Perceiving Subtle Sexism: Mapping the Social-Psychological Forces and Legal Narratives That Obscure Gender Bias PERCEIVING SUBTLE SEXISM: MAPPING THE SOCIAL-PSYCHOLOGICAL FORCES AND LEGAL NARRATIVES THAT OBSCURE GENDER BIAS DEBORAH L. BRAKE* In early January of 2007, the AALS Section on Women in Legal Education held a panel discussion on "Subtle Sexism in Our Everyday Lives" at the AALS Annual meeting in Washington, D.C. Such discussions about the barriers facing women in the legal profession often trigger a fatigue with talking about gender and a denial by some that gender remains worthy of attention.1 The denial of gender bias can occur at a collective level, in which detractors urge "getting past" gender in setting an agenda, and at an individual level, in which individuals deny the role of gender bias (or gender privilege) in their own lives.2 The denial of gender bias at the * Professor of Law, University of Pittsburgh School of Law. A version of this paper was first presented at the AALS Annual Meeting at a panel on "Subtle Sexism in Our Everyday Lives," sponsored by the Section on Women in Legal Education. Many thanks to the participants at that program, and especially to my colleague Pat Chew for organizing the program and for reading and commenting on a draft of this Article. I benefited from the research assistance of Christopher Helms on this and related projects. 1 See, e.g., Deborah L. Rhode, Midcourse Corrections: Women in Legal Education, 53 J. LEGAL EDUC. 475, 476 (2003) (discussing the conventional view of legal education that "'the woman problem' has been solved and equal opportunity is an accomplished fact," and the reality that "our partial progress has itself become an obstacle to further change"); Judith Reskin, A Continuous Body: Ongoing Conversations About Women and Legal Education, 53 J. LEGAL EDUC. 564, 568 (2003) (discussing skeptical and less friendly reactions to ongoing conversations about gender in the legal academy, in which skeptics urge participants to "move on" to other topics and critics even "argu[e] that, by calling attention to problems of inequality and by acting affirmatively to remedy them, we create inequality"). Participants in such conversations often feel compelled to justify the ongoing significance of gender, acknowledging progress but pointing to the half-empty glass. See, e.g., Deborah Jones Merritt & Barbara F. Reskin, New Directions for Women in the Legal Academy, 53 J. LEGAL EDUC. 489 (2003) (discussing the progress women have made in the legal academy since 1970, but also the persistence of disadvantages that block the path to equality for women in the legal academy, and women of color in particular). 2 See Sylvia A. Law, Good Intentions Are Not Enough: An Agenda on Gender for Law School Deans, 77 IowA L. REv. 79, 81 (1991) ("[I]t is still quite common in American 680 Columbia Journal of Gender and Law [Vol. 16:3 individual level can fuel a collective denial of the importance of gender issues. This Article explores the complexity of perceiving gender bias at the individual level, which in turn affects understandings of the role of gender in society more broadly, and surveys some of the psychological processes that contribute to the denial of gender bias in everyday life. In this Article, I am concerned both with how women law faculty and women lawyers-the immediate subjects of the panel discussion-perceive gender bias, and with the more general forces that complicate perceptions of gender bias across professional lines. Sorting out the influence of gender on an individual's professional life is no easy task. For women law faculty, contemplating the role of subtle sexism might raise a nagging set of questions. How does my gender affect my professional life? Do students react differently to me because of it?3 What explains that small but disturbing set of hostile course evaluations in my large required classes, and do my male colleagues get them too?4 Is it my imagination, or are those guys in the back row challenging my authority? 5 Is the tenure process gendered or just crazy? 6 How does gender legal education to hear comments that seem to deny gender discrimination is deeply entrenched and takes other than intentional forms."). 3 See Christine Haight Farley, Confronting Expectations: Women in the Legal Academy, 8 YALE J. L. & FEMINISM 333, 336 (1996) (citing results of a study showing that "forty-eight percent of all women students and seventy-three percent of minority women students believe that female professors, more than male professors, must prove their competence to their students"). 4 See infra note 172 for sources suggesting that gender and race bias influence student evaluations. See also Joan M. Krauskopf, Touching the Elephant: Perceptions of Gender Issues in Nine Law Schools, 44 J. LEGAL EDUC. 311, 315 (1994) (discussing the "hostility or a presumption of incompetence from students, especially male students" toward women law teachers). 5 See, e.g., Rhode, supra note 1, at 486 ("Women faculty, particularly women of color, often experience classroom challenges to their competence and authority."); Farley, supra note 3, at 334, 342 (stating that "[b]ecause women lack the presumption of competence, they are continuously being challenged, resulting in a hostile 'prove it' atmosphere," and noting that "often women will spend much more time on class preparation in order to anticipate every possible line of attack they may face in class"). 6 As Tina Grillo has remarked: [T]he system of getting tenure is crazy for everybody .... The problem that faces minority women is this: we cannot tell how much of the craziness has to do with sexism and racism-alone or in combination, expressed overtly or expressed covertly through institutional politics- and how much has to do with a fundamentally crazy system. Perceiving Subtle Sexism shape my professional obligations and institutional commitments? 7 Why are so many women on the faculty silent during faculty meetings, while almost all of the big talkers are men? 8 These questions defy easy and absolute answers, and in this respect, are no different than similar questions that might interrogate the role of gender in other professional settings. We live in a world where gender, race, and sexuality form a complex web of identity that subtly affects, in myriad ways, how people respond to us. As we know from life experience and the study of law, facts are messy and causation is tricky to pin down. How do we make sense of the uncertainty? Social psychologists use the term "attributional ambiguity" to refer to the "uncertainty about whether the outcomes you receive are indicators of something about you as an individual, or indicators of social prejudices that other people have against you because of your stigma." 9 Short of written policies that openly discriminate or overt expressions of prejudice, discerning the presence of gender bias necessarily involves attributional ambiguity. The messy reality of perceiving gender bias contrasts sharply with the common assumption, reflected in discrimination law, that a person's belief that she has experienced discrimination is fixed and immediate. My interest in exploring the process by which people come to believe they have experienced gender bias grew out of my work on an amicus curiae brief Tina Grillo, Tenure and Minority Women Law Professors: Separating the Strands, 31 U.S.F. L. REV. 747, 747-48 (1997). 7 See Susan B. Apel, Gender and Invisible Work: Musings of a Woman Law Professor, 31 U.S.F. L. REV. 993 (1997) (discussing the "invisible work" disproportionately done by women law professors, including "community building," student counseling, and providing support and service to colleagues and the institution). 8 See Martha Chamallas, The Shadow of Professor Kingsfield." Contemporary Dilemmas Facing Women Law Professors, 11 WM & MARY J. WOMEN & L. 195, 204-06 (2005) [hereinafter Chamallas, The Shadow of Professor Kingsfield] (discussing the subtle loss of status that occurs when women's expressed ideas are not valued, and explaining how this process causes even senior and accomplished women faculty to be less outspoken at their home institutions); Merritt & Reskin, supra note 1, at 493 (discussing the institutional messages that leave "women less sure of their voices" and "less likely to speak in the faculty lounge and slower to commit their ideas to paper"). See also Martha Albertson Fineman, The New "Tokenism," 23 VT. L. REV. 289 (1998) (explaining why the increasing numbers of women in the legal academy have not significantly challenged the norms and culture of the legal academy). 9 Brenda Major et al., Attributions to Discrimination and Self-Esteem: Impact of Group Identification and Situational Ambiguity, 39 J. EXPERIMENTAL SOC. PSYCHOL. 220, 220 (2002) [hereinafter Major et al., Attributions to Discrimination]. 2007] Columbia Journal of Gender and Law filed in Ledbetter v. Goodyear Tire & Rubber Co.'0 In Ledbetter, the Eleventh Circuit ruled that an employee must bring a Title VII pay discrimination claim within 180 days of the time the intentionally discriminatory decision was first made or else be time-barred from ever bringing the claim, even if the employee continues to receive less pay because of sex." The lower court's decision treats pay discrimination as analogous to other "discrete" discriminatory acts governed by the rule the Supreme Court adopted in National Railroad Passenger Corporation v. Morgan.l2 In Morgan, the Court held that each discriminatory act, such as a hiring, firing, promotion, demotion, or transfer decision, triggers Title VII's statute of limitations period, even if it is part of a related pattern of discrimination that extends beyond that "discrete" act.' 3 In so holding, the Court rejected the more lenient continuing violation theory, which lower courts had applied to allow plaintiffs to toll the limitations period for discriminatory acts that are part of a larger pattern of related discrimination. 14 Without directly stating where pay discrimination claims fall on this continuum, the Court in Morgan distinguished discrete discriminatory acts from hostile environment harassment. 5 The Court crafted a different and more lenient rule for hostile environment harassment, treating each harassing act as part and parcel of the larger pattern of harassment, and tolling the running of the limitations period until the last act of harassment occurs.16 The Court justified the special rule for harassment because it typically requires a number of harassing acts in order 10 See Brief for National Partnership for Women & Families et al. as Amici Curiae Supporting Petitioners, Ledbetter v. Goodyear Tire & Rubber Co., Inc., 127 S. Ct. 2162 (2007) (No. 05-1074). As this issue was going to press, on May 29, 2007, the Supreme Court handed down its decision in Ledbetter v. Goodyear Tire & Rubber Co., Inc., 127 S. Ct. 2162 (2007). A 5-4 majority ruled that a Title VII pay discrimination claim's filing period begins running at the time the intentionally discriminatory pay decision was made, and any ongoing pay disparities resulting from that decision are time-barred once the 180-day filing period expires. " 412 F.3d 1169 (11 th Cir. 2005). 12 536 U.S. 101 (2002). 131d. at 110-11. 14 Id. at 108, 113-14. " Id. at 115. 16Id. at 117. [Vol. 16:3 Perceiving Subtle Sexism to have an actionable claim. 17 At no point did the Court consider how long people take to realize they have experienced discrimination, nor did it acknowledge the difficulty of discerning discrimination. This body of precedent assumes that employees possess immediate and certain knowledge of the moment in time at which discrimination occurs. Perceiving discrimination is assumed to be uncomplicated, such that plaintiffs who do not complain shortly after discrimination occurs are vulnerable to criticism for "sitting on their rights."'18 One modest exception in the doctrine moderates this assumption, but it nevertheless stops short of acknowledging the complexity involved in perceiving discrimination. Existing Title VII law leaves open the possibility that a discovery rule might delay the running of the limitations period until the plaintiff knew or should have known that she experienced discrimination. This concession, however, has not seriously disrupted the law's assumption that knowledge of discrimination is unproblematic. For one thing, the Supreme Court has been content to leave the existence of a discovery rule in Title VII cases an open question, an indication that it views justifiable delays in perceiving discrimination to be the exception, rather than the norm. 19 In like fashion, the Eleventh Circuit in Ledbetter simply observed that the question of whether equitable considerations justified a tolling of the limitations period in that particular case had not been litigated. z More importantly, perhaps, even those lower courts that have adopted a discovery rule in Title VII cases have failed to grapple with the complexity of perceiving discrimination. These courts have applied the "7 Id. at 117-18. 18 For example, in a case where an employer counted time off for pregnancy against employees in calculating their service credit (and thus seniority and eligibility for early retirement), the employer argued that Title VII discrimination charges filed many years after the Pregnancy Discrimination Act's passage were untimely, and that a district court's decision to that effect should be upheld. See Final Brief of Defendant-Appellee at 7, EEOC v. Ameritech Servs., Inc., 129 F. App'x 953 (6th Cir. 2005) (No. 04-3496) ("The court relied on [the Sixth Circuit's] precedents, which preclude employees from sitting on their rights and imposing 'an open-ended period of liability for the employer."'). 19 National RR Passenger Corp. v. Morgan, 536 U.S. 101, 105, 114 n.7 (2002) (stating that equitable tolling principles apply under Title VII, but not discussing the applicability of a discovery rule); id. at 124 (O'Connor, J., concurring) (expressing her belief that "some version of the discovery rule applies to discrete-act claims"). See also id at 113 (cautioning that equitable doctrines such as tolling and estoppel "are to be applied sparingly"). 2 412 F.3d 1169, 1180 n.16 (1lth Cir. 2005). 2007] 683 Columbia Journal of Gender and Law [Vol. 16:3 discovery rule to set the moment in time when the plaintiff should have known of the alleged discrimination at the point when the plaintiff first learned of the adverse job decision (or in the case of pay, that a male comparator earns more), rather than the moment when the plaintiff actually perceived the discrimination. 2 1 Thus, even the law's allowance for a discovery rule does not seriously disrupt the assumption that knowledge of bias is uncomplicated and readily inferable from the existence of an injury. In addition to the assumptions reflected in law, the assumptions embedded in popular culture also conflict with the realities of how people perceive discrimination. The dominant story in mainstream culture is that women and minorities are hyper-vigilant in perceiving bias, to the point of mistakenly perceiving sexism and racism when it does not really exist. Mainstream culture is replete with derogatory references to "feminazi" women who blame everything on gender, and with depictions of strident women who are too quick to blame sexism for their troubles.22 Likewise, people of color are derided for "playing the race card," as if an attribution to race is always an artifice and never an insight.2 3 This cultural narrative 21 See, e.g., Inglis v. Buena Vista Univ., 235 F. Supp. 2d 1009, 1025 (N.D. Iowa 2002); Oshiver v. Levin, 38 F.3d 1380, 1386-87 (3d Cir. 1994). But see Hamilton v. First Source Bank, 928 F.2d 86, 90 (4th Cir. 1990) (en banc) (rejecting a discovery rule for pay claims under the Age Discrimination in Employment Act and holding that the "last possible time that pay discrimination could have occurred was the date when [the plaintiff] received his final paycheck"). 22 See, e.g., Alicia Mundy, "Women 's Rights" Getting Left Out, SEATrLE TIMES, May 13, 2006, at A6 (discussing the reluctance of nationally prominent women politicians to publicly discuss "women's rights," lest they be labeled "feminazis" by talk radio hosts); Michelle Garcia, New York Women in Line to Get Restroom Relief, WASH. POST, May 27, 2005, at A3 (reporting that a proponent of city legislation to increase restroom facilities for women was labeled a "feminazi"); Stephen Kiehl, Eatery Draws Fire for "Sexy Nurses": Arizona's Heart Attack Grill Asked to Stop Using Servers' Costume, L.A. TIMES, Nov. 24, 2006, at A31 (describing the hostile reaction to complaints about sexually objectifying uniforms for servers at one restaurant, including labeling the complainant a "feminazi"); Mike Rosen, Editorial, AAUW's Sexual Fantasies, ROCKY MTN. NEwS, Feb. 10, 2006, at 43A (criticizing an American Association of University Women study finding high rates of sexual harassment on college campuses and stating that "hypersensitive women or outright man-haters have been led to believe they have some right to be insulated from anything, real or imagined, to which they might conceivably take offense"). 23 A LexisNexis search for the term "playing the race card" in the most recent two years of news (search date, Mar. 6, 2007) turned up 1151 stories including that term. Typically, the term is used to criticize persons for cynically and strategically inserting race into an area of public discourse where it does not belong. See, e.g., Harold Johnson & Timothy Sandefur, Editorial, Judge Doesn't Deserve the Dixiecrat Treatment, S.F. CHRON., May 2, 2005, at B5 ("To be sure, many of [Judge Janice Rogers] Brown's conservative supporters can be accused of playing the race card."); Willis Shalita, Editorial, Black 2007] Perceiving Subtle Sexism 685 encourages suspicious treatment of those who would attribute adverse outcomes to gender or race bias.24 Research in the field of social psychology suggests that the realities of perceiving bias are much more complex than either legal doctrine or dominant cultural understandings acknowledge. First, the widespread cultural assumption of hyper-vigilance is largely a myth. Although there is modest evidence that some persons who belong to stigmatized social groups are highly sensitive to prejudice cues, 5 the weight of evidence suggests that under-perception of gender bias is closer to the norm than hyper-vigilance.26 Studies consistently show, for example, that the vast majority of women who experience behavior that objectively qualifies as sexual harassment do not perceive that they have been sexually harassed.27 More generally, social American From Africa Offers His View on Obama's "Blackness," S.F. CHRON., Mar. 1, 2007, at B7 ("Black America better wake up and smell the coffee. Time for playing the race card in national politics is over, but much more than that, it is a disgrace, especially coming from victims of racism."). 24 Cf Rhode, supra note 1, at 487 ("Women students who express strong feminist views have been stigmatized for 'overreaction' or for behaving like 'feminazis' or 'manhaters."'). 25 See, e.g., Cheryl R. Kaiser & Brenda Major, A Social Psychological Perspective on Perceiving and Reporting Discrimination, 31 L. & SOC. INQUIRY 801, 803-04 (2006) (summarizing and describing as "sparse" the "empirical evidence that members of historically disadvantaged groups claim discrimination when none exists, or even that they are especially sensitive to and vigilant for discrimination"). 26 See id. at 804-06 (summarizing studies supporting the view that people err on the side of denying or minimizing discrimination targeting them); Brenda Major & Cheryl R. Kaiser, Perceiving and Claiming Discrimination, in HANDBOOK OF EMPLOYMENT DISCRIMINATION RESEARCH: RIGHTS AND REALITIES 285, 286-87 (Laura Beth Nielsen & Robert L. Nelson eds., 2005) (citing sources supporting the view that "members of disadvantaged groups typically miss, underestimate, or deny the extent to which they are personally targets of prejudice"); Elizabeth H. Dodd et al., Respected or Rejected: Perceptions of Women Who Confront Sexist Remarks, 45 SEx ROLES 567, 568-69 (2001) [hereinafter Dodd et al., Respected or Rejected] (summarizing research showing that women tend to explain away sexism, despite evidence that it has occurred); Charles Stangor et al., Reporting Discrimination in Public and Private Contexts, 82 J. PERSONALITY & SOC. PSYCHOL. 69, 69 (2002) ("[P]rior research has shown that members of stigmatized groups are in many cases unlikely to report that negative events that occur to them are due to discrimination, even when this is a valid attribution for the event."). 27 Vicki J. Magley et al., Outcomes of Self-Labeling Sexual Harassment, 84 J. APPLIED PSYCHOL. 390 (1999) [hereinafter Magley et al., Outcomes of Self-Labeling] (citing and discussing research documenting a wide disparity between the numbers of women who experience unwelcome, offensive sexual misconduct and those who identify their experiences as sexual harassment). See also Beth A. Quinn, The Paradox of Complaining: 686 Columbia Journal of Gender and Law [Vol. 16:3 psychologists have long observed a disconnect between women's recognition that women as a group experience discrimination and individual women's widespread denial that they have personally experienced it. 28 This finding was first noted in a 1978 study in which 400 male and female workers rated their personal job satisfaction and grievances no differently, despite objective evidence that the women in the study were subjected to sex discrimination.2 9 Subsequent research has updated and added to the evidence supporting the under-perception thesis. 30 Second, social psychology research refutes the assumption that knowledge of discrimination is uncomplicated and immediate. In reality, Law, Humor, and Harassment in the Everyday Work World, 25 L. & SoC. INQUIRY 1151, 1156 (2000) [hereinafter Quinn, The Paradox of Complaining] (rejecting a lack of legal understanding as a sufficient explanation for women's resistance to labeling their experiences as sexual harassment). 28 See Faye Crosby, The Denial of Personal Discrimination, 27 AM. BEHAV. SCI. 371 (1984) [hereinafter Crosby, The Denial of Personal Discrimination]. As James M. Olson and Carolyn L. Hafer observe: "Of course, it is a statistical impossibility for all members of a group to experience less discrimination than other members." James M. Olson & Carolyn L. Hafer, Tolerance of Personal Deprivation, in THE PSYCHOLOGY OF LEGITIMACY: EMERGING PERSPECTIVES ON IDEOLOGY, JUSTICE, AND INTERGROUP RELATIONS 157, 163 (John T. Jost & Brenda Major eds., 2001) [hereinafter Olson & Hafer, Tolerance of Personal Deprivation]. These authors also note that the discrepancy between perceptions of discrimination against one's social group and perceptions of discrimination against oneself hold true for a wide variety of disadvantaged groups and "crosses racial, gender, and economic boundaries." Id. at 164; see also John T. Jost, Negative Illusions: Conceptual Clarification and Psychological Evidence Concerning False Consciousness, 16 POL. PSYCHOL. 397, 404-05 (1995) [hereinafter Jost, Negative Illusions]. These findings are consistent with the tendency of minority group members to accommodate and internalize the perspective of the majority. 29 Crosby, The Denial of Personal Discrimination, supra note 28, at 372-73. See also Donald M. Taylor et al., The Personal/Group Discrimination Discrepancy: Perceiving My Group, But Not Myself To Be a Target for Discrimination, 16 PERSONALITY & SOC. PSYCHOL. BULL. 254 (1990). 30 See, e.g., Jacquie D. Vorauer & Sandra M. Kumhyr, Is this About You or Me? Self- Versus Other-Directed Judgments and Feelings in Response to Intergroup Interaction, 27 PERSONALITY & SOC. PSYCHOL. BULL. 706 (2001) (reporting results of a study in which members of a racially stigmatized group who interacted with a prejudiced peer felt badly after the interaction, but interpreted the negative feelings provoked by the interaction as having to do with themselves rather than the other person's prejudice); see also Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 163 (citing research finding that "individuals who might be expected to report a lot of discrimination often report very little"). For findings on the denial of personal disadvantage by women in the legal profession, see Faye Crosby et al., Discontent Among Male Lawyers, Female Lawyers, and Female Legal Secretaries, 13 J. APPLIED SOC. PSYCHOL. 183 (1983). Perceiving Subtle Sexism there are a number of obstacles to perceiving oneself as a victim of discrimination. These obstacles function as a form of "static" or "interference" that clouds the picture and distorts the process of perception. Such "interference" is especially likely when gender bias takes a subtle form, without obvious markers. Not surprisingly, people are much better at detecting prejudice when the intensity of prejudice cues increases, and are less able to do so when the bias is subtle or more ambiguous. 3' In a world characterized by attributional uncertainty, a number of psychological and social processes converge to create a good deal of interference with perceptions of gender bias. These processes make "seeing" how sexism affects daily life a complicated endeavor. Rather than clear the fog to get a better picture of how gender influences life experience, the dominant narratives of discrimination law tend to reinforce the perceptual static. This is perhaps especially true for those of us in the legal academy who are more likely to have our understandings and perceptions shaped by the discourses and narratives of law. This Article surveys the social psychology literature on the processes that complicate perceptions of gender bias and explores how these processes are reinforced by the narratives of discrimination law. Part I draws on social psychology research to chart the obstacles that interfere with perceiving subtle sexism. Part II considers how the narratives of discrimination law further suppress perceptions of bias. The point is not that law leaves behind subtle gender bias in the discarded category of "not actionable"-an argument well made by other legal scholars.32 Rather, it is that law's knowledge-producing function obscures subtle gender bias from view, reinforcing the ideology that merit, and not gender, explains women's situations in the workplace. As explained below, law performs this function by setting up narratives that constrain the likelihood of perceiving and challenging gender bias. Finally, Part III briefly considers the example of student evaluations as an illustration of how these processes can converge to muddy insights into how and whether gender bias shapes our professional lives. 3 Brenda Major et al., Prejudice and Self-Esteem: A Transactional Model, 14 EUR. REV. SOC. PSYCHOL. 77, 81 (2003) [hereinafter Major et al., Prejudice and Self-Esteem]. 32 See, e.g., Martha Chamallas, Deepening the Legal Understanding of Bias: On Devaluation and Biased Prototypes, 74 S. CAL. L. REV. 747, 747-53 (2001) [hereinafter Chamallas, Deepening the Legal Understanding] (discussing the limits of modem discrimination doctrine and citing legal scholarship critiquing those limits). 2007] 687 Columbia Journal of Gender and Law I. LESSONS FROM SOCIAL PSYCHOLOGY ON THE COMPLEXITY OF PERCEIVING GENDER BIAS A number of cognitive and motivational processes exist that make it difficult for a person to see herself as the target of gender bias. This section surveys some of these barriers and explores how they intersect to submerge and complicate individual awareness of gender bias and its influence on women's lives. A. The Belief in a "Just World," the Ideology of Individual Responsibility, and the Reluctance to Blame Others Social psychologists have identified a psychological drive to believe that the world is fundamentally "just" in the sense of meritocratic, such that people get what they deserve and are not held back by unfair considerations such as gender or race.33 Attribution of negative outcomes to discrimination clashes with this belief in a "just world." As a result, women and people of color who strongly believe in a "just world" tend to discount discrimination as an explanation for negative outcomes and instead blame themselves.34 For example, researchers have found that the more that members of lower-status groups (specifically, in this research, Latino/a Americans and women) "endorsed the ideology of individual mobility (e.g., agreed with statements such as "advancement in American society is possible for individuals of all ethnic groups"), the less likely they were to report that they personally, or members of their group, were targets of discrimination. 35 The influence of "just world" ideology on perceptions of discrimination is heightened when, as is often the case, there is ambiguity 33 See, e.g., Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 159-63. As originally conceived, the "just world" theory had its primary analytical force in explaining peoples' reaction to the suffering of others. Later work has demonstrated the force of the theory in also explaining how people make sense of their own suffering. Id. at 159-60. 34 Id. at 161. Because the belief in a "just world" tends to suppress perceptions of discrimination both at the collective and the individual level, it does not explain the disconnect noted above between perceiving discrimination against one's social group while denying any personal experience with discrimination. On that score, see infra notes 46-56 and accompanying text (discussing the belief in individual responsibility and the reluctance to blame others). 35 Major et al., Prejudice and Self-Esteem, supra note 31, at 82. The authors add: "They were also less likely to blame discrimination when a higher-status confederate (European American; man) rejected them for a desirable role." Id. [Vol. 16:3 2007] Perceiving Subtle Sexism 689 about why a particular negative outcome occurred.36 When prejudices are subtle and circumstances ambiguous, adherence to "just world" ideology is especially likely to lead members of stigmatized groups to favor internal explanations over bias as the reason for a poor outcome.37 Interestingly, adherence to "just world" ideology does not have the same effect on perceptions of discrimination for all social groups. Members of high-status social groups, such as white males, are more likely to attribute a negative outcome to personal discrimination when they adhere to the belief in a "just world., 38 Accordingly, experimental research has found that while adherence to a belief in a "just world" lessens the likelihood that members of lower-status groups will perceive discrimination when rejected for a position by higher-status group members, members of higher-status groups who endorse "just world" ideology are more likely to blame 39discrimination for their rejection by a member of a lower-status group. 36 Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 163. 37 Major et al., Attributions to Discrimination, supra note 9, at 230 ("[A]mbiguous situations appear to be especially difficult for members of stigmatized groups. Because they disguise prejudice, they create uncertainty and interfere with the target's ability to discount their own role in producing negative outcomes."). 38 See Kaiser & Major, supra note 25, at 812; Major et al., Prejudice and Self Esteem, supra note 31, at 82. 39 As Kaiser and Major explain, Because endorsing this meritocratic worldview results in seeing low- status group members as deserving of their poor outcomes, the more low-status group members endorse these beliefs, the more they will minimize the extent to which they face discrimination .... In contrast, because endorsing the meritocratic worldview leaves members of high- status groups feeling entitled to their privileged position, the more they endorse the worldview, the more sensitive they will be towards perceiving signs of reverse discrimination. In other words, endorsing the meritocratic worldview leads members of high-status groups to anticipate preferential treatment (because they assume they have greater abilities), and they will feel threatened and slighted when members of low-status groups receive better treatment than they do . . . . Hence, seeing their negative outcomes as stemming from reverse discrimination can be one way to maintain their faith in the worldview that conveys that they should be at the top of the social hierarchy. Kaiser & Major, supra note 25, at 808-09 (internal citations omitted). 690 Columbia Journal of Gender and Law [Vol. 16:3 Because the belief in a "just world" is pervasive, this ideology exerts a strong influence on perceptions of discrimination. 40 It may be especially powerful within institutions highly steeped in the ideology of merit, such as the legal profession in general and the legal academy in particular.4 ' Settings where the belief in meritocracy is especially pronounced discourage perceptions of bias against those who do not rise to the top. As defined above, the belief in a "just world" is infused with the ideology of individual responsibility. This ideology appeals to many people because it enables them to believe that they have control over their destiny.42 Seeing oneself as a victim of discrimination contradicts that belief. The ideology of individual responsibility "turn[s] the word victim into a synonym for failure or irresponsibility.'43 This belief system creates an aversion to being perceived as a victim of discrimination, especially when one's victim status is linked to membership in a social group whose members are stigmatized and devalued.44 The belief in individual responsibility enables a person to acknowledge that members of her social group might experience discrimination, yet preserves her individual agency by discounting the possibility that she herself has experienced discrimination.45 For example, 40 See id. at 806-08 (describing "the meritocratic worldview" and its prevalence in mainstream United States culture). " See id at 810-12 (discussing research finding that low-status groups' attributions to discrimination decrease when targets are first primed with messages highlighting a meritocratic worldview, and concluding that "environments where the meritocratic worldview is pervasive (such as meritocracy-based employment institutions), might lead members of devalued groups to minimize discrimination and members of high-status groups to become more sensitive to signs of reverse discrimination"). See also Martha S. West, Gender Bias in Academic Robes: The Law's Failure to Protect Women Faculty, 67 TEMP. L. REv. 67, 139-43 (1994) (discussing the prevalence of the "myth of meritocracy" in the academy and how it operates to justify the unequal position of women faculty). 42 Kaiser & Major, supra note 25, at 808 (discussing the psychological benefits derived from endorsement of the meritocratic worldview, including a sense of control over one's destiny). 43 Magley et al., Outcomes of Self Labeling, supra note 27, at 392. 44 Id. at 392-93; Carolyn L. Hafer & James M. Olson, Beliefs in a Just World, Discontent, and Assertive Actions by Working Women, 19 PERSONALITY. & SOC. PSYCHOL. BULL. 30, 35 (1993) (explaining that people who hold strong beliefs in a just world tend to minimize discrimination and blame themselves for poor outcomes). 45 See supra note 42. Perceiving Subtle Sexism women who are sexually harassed might resist labeling their experience as such to preserve their belief in individual responsibility and avoid the stigmatized identity of a harassment victim. 46 Likewise, women's generally low level of apparent dissatisfaction with their lower pay-a phenomenon that is described below47 -may reflect their aversion to being cast as victims if they attribute the disparity to discrimination. 48 In my experiences representing young women in Title IX challenges to unequal athletic opportunity, I was often struck by how frequently female athletes in these cases went out of their way to proclaim that they were not feminists.49 In their unprompted disclaimers, I sensed an anxiety about being seen as victims for bringing a discrimination suit, and a desire to avoid a deeper stigma of victimhood that a broader challenge to sexism might entail. Even some young women who identify as feminists emphatically deny that they personally have experienced sexism-a phenomenon I have often witnessed among students in my gender law classes. Some social psychologists suggest that the desire to see oneself as an agent with control over one's life, rather than as a victim of external forces, best explains the gap between the perception of widespread discrimination against one's social group and the widespread denial of personal discrimination.5 ° A corollary to the belief in individual responsibility is the reluctance to blame another person for a negative outcome in one's own 46 See Quinn, The Paradox of Complaining, supra note 27, at 1173 (explaining that complaining of sexual harassment saddles the complainant with a "stigmatized" identity, and quoting one manager as stating that "making a claim of sexual harassment 'is sort of like rape, it tends to reflect as badly on the person filing the report as it does the person being accused'); Adrienne D. Davis & Stephanie M. Wildman, The Legacy of Doubt: Treatment of Sex and Race in the Hill-Thomas Hearings, 65 S. CAL. L. REV. 1367, 1374-75 (1992) (discussing the "blame the victim" attitude that confronts women who are sexually harassed). 47 See infra notes 68-70 and accompanying text. 48 Cf Brenda Major, From Social Inequality to Personal Entitlement: The Role of Social Comparisons, Legitimacy Appraisals, and Group Membership, 26 ADVANCES EXPERIMENTAL PSYCHOL. 293, 325-26 (1994) [hereinafter Major, From Social Inequality] (noting that when people are aware that women receive less pay than men, they tend to infer that differences in individual marketability, including such factors as commitment to the workforce and differences in job responsibilities, job performance, or job qualifications, explain the pay disparity, even when they do not). 49 From 1992-1998, 1 was a staff attorney and then senior counsel at the National Women's Law Center and was involved in litigating Title IX challenges to sex discrimination in interscholastic and intercollegiate athletics. 50 Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 164. 2007] 692 Columbia Journal of Gender and Law [Vol. 16:3 life. Under common understanding and prevailing legal doctrine, if there is sexism there must be a sexist. 51 Yet, social psychologists have observed an emotional reluctance to perceive discrimination against a person when it requires identifying an individual villain. 52 Termed "blame avoidance," this phenomenon helps explain why women and other members of subordinated social groups are more likely to recognize the existence of discrimination against their social group but deny that they have personally experienced it.53 Recognizing systemic and anonymous discrimination against women does not require an identifiable villain, but perceiving discrimination against an individual woman does. One study in support of the finding that people resist blaming others under conditions suggestive of prejudice involved pairs of Canadian students consisting of one white student and one aboriginal student who were grouped together for the purpose of interacting in a get-acquainted conversation. 54 Prior to the pairing, the white students were given a test that measured racial prejudice. The students were then grouped into mixed-race pairs for discussions. The researchers found that, in pairings with a highly- prejudiced white student, the aboriginal student left the interaction feeling worse than the aboriginal students who were paired with less-prejudiced whites, suggesting that the former set of interactions went badly.55 However, instead of blaming their white cohorts or perceiving them as prejudiced, the aboriginal students who were paired with the high-prejudice 51 But see Catharine MacKinnon, Difference and Dominance: On Sex Discrimination, in FEMINISM UNMODIFIED: DISCOURSES ON LIFE AND LAW 43-45 (1987) (criticizing conventional discourse about sex discrimination for seeing gender inequality as a problem of sexism). MacKinnon states: If gender were merely a question of difference, sex inequality would be a problem of mere sexism, of mistaken differentiation, of inaccurate categorization of individuals. . . . But if gender is an inequality first, constructed as a socially relevant differentiation in order to keep that inequality in place, then sex inequality questions are questions of systemic dominance, of male supremacy, which is not at all abstract and is anything but a mistake. Id. 52 See Crosby, The Denial of Personal Discrimination, supra note 28. " See id 54 See Vorauer & Kumhyr, supra note 30. " Id. at 716. Perceiving Subtle Sexism white students blamed themselves for their negative feelings.56 From this and other research, the authors surmised that people tend to think about interactions with others in terms of what they reveal about themselves and not what they reveal about the others. The authors interpreted their findings to support the thesis that in everyday life members of low-status groups are unlikely to detect prejudice, and instead internalize the negative emotions produced by interactions that are tainted by prejudice. The belief in a "just world," the belief in individual responsibility, and the reluctance to blame others function together to discourage perceptions of gender bias. B. The Default Preference for Within-Group Comparisons and the Effect on Individual Sense of Entitlement Another factor that influences perceptions of bias is a person's sense of entitlement. An individual's belief that she has experienced discrimination is contingent upon her belief in the illegitimacy of her current treatment. As social psychologist Brenda Major explains, "beliefs about entitlement are a critical determinant of how members of social groups react affectively, evaluatively, and behaviorally to their socially distributed outcomes. 57 People are not likely to perceive bias or prejudice unless they believe they are entitled to better treatment. Expectations about the legitimacy of outcomes and one's sense of entitlement are thus critical factors shaping the perception of bias. These beliefs and expectations, in turn, are largely driven by the process of social comparison.58 Information about what level of treatment others receive enables people to form an opinion about what level of treatment they deserve. Comparison to others provides necessary benchmarks about what outcomes are possible and justifiable. The selection of comparators is therefore critical in shaping perceptions of fairness and bias. Social scientists have found that people tend to compare themselves to members of the same social group. In particular, working women tend to compare their situations to those of other working women because of their 56 Id. 57 See Major, From Social Inequality, supra note 48, at 293-94. 58 See id. at 298-300 (explaining that feelings of entitlement and deservingness shape expectations and social justice perceptions, and that the process of comparison to others is critical in determining beliefs about entitlement). 2007] 693 Columbia Journal of Gender and Law perceived similarity, proximity, and ease of comparison with one another.59 Structural features of the workplace, including vertical and horizontal gender segregation, reinforce within-gender comparisons by bolstering the similarity and proximity of women in the workforce. 60 The under-valuation of women's work and the under-payment of women further reinforce a sense of similarity among women for comparison purposes, as does women's disproportionate share of home responsibilities. 6' At the same time, women's different situations from men on these scores lessen the salience of comparisons to men.6 2 Within-gender comparisons, while understandable, function to legitimize and obscure gender differences in treatment. Working women who compare themselves to other working women develop a diminished sense of entitlement.63 As a result, the very existence of systematic discrimination against women as a group makes individual instances of bias harder to detect because the group experience affects the level of treatment individual women perceive as normal and legitimate.64 At the same time, men tend to compare themselves to other men, leading them to expect the continuation of the privileges that correspond to their gender.65 A similar effect on women's sense of entitlement follows from the related tendency of individuals to compare their current situations with their 59 See Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 166- 67; Major, From Social Inequality, supra note 48, at 302. 60 See Major, From Social Inequality, supra note 48, at 314. 61 Id. at 314-15. 62 Id. at 315. See also id. at 303 ("[P]eople tend to make intragroup rather than intergroup comparisons when estimating what they deserve because of the greater availability, and assumed greater similarity and diagnosticity, of the former."). 63 See id. at 320-21 (explaining that women's default within-group comparison reference point leads to lower expectations for pay than men have). 64 See id. at 294. 65 See id. at 321-22. Major notes: Women and men estimate their personal deserving against a (same) sex- stereotyped judgment standard. . . . Because women and people doing 'women's jobs' are typically paid less than men and people doing 'men's jobs,' women estimate their personal deserving and evaluate their outcomes against a lower reference standard for pay than do men. 694 [Vol. 16:3 Perceiving Subtle Sexism past experiences. A person's past experience shapes her expectations and sense of legitimacy with respect to her present treatment. This too tends to legitimize and render invisible systemic discrimination. For example, a woman's prior pay has a strong influence on her expectations about the level of pay she currently deserves. 66 Since women on average receive less pay than men, the comparison to past experience is likely to suppress a woman's sense of entitlement with respect to current pay. In this way, individual discrimination becomes self-reinforcing: people who have experienced discrimination believe that they deserve less and do not perceive persisting disadvantage as discriminatory or illegitimate, while persons with privilege have a strong sense of entitlement to continued favorable treatment. 67 Together, gender-specific comparisons and comparison to one's prior experience obscure systemic bias and legitimize the status quo. 68 These comparison processes help explain women's suppressed sense of entitlement, which social scientists have amply documented. In one study in which women and men were asked to set their pay for performing a specified task without information on expected or baseline rates, women paid themselves only sixty-one percent of what men paid themselves.6 9 In a follow-up study in which men and women were paid the same amount to perform a task but were told to work as long as they believed appropriate, the women worked one-third longer than the men in the study. 70 The evidence suggests that the use of within-gender comparison standards is "by 66 See id. at 322. The potency of the comparison to one's own past experience may even overcome a bias toward within-group comparisons. For example, when well-paid women were reminded of their prior pay, which was equal to men's, they paid themselves equally to the men when given the task of setting their pay for a specific work task. Id. 67 See id. at 307-08, 321-22. 68 See id. at 303 ("[P]eople typically feel they deserve the same treatment or outcomes that they have received in the past or that others like themselves receive."). 69 See Jost, Negative Illusions, supra note 28, at 404. See also Major, From Social Inequality, supra note 48, at 313-14 (discussing research explaining women's paradoxical satisfaction with their pay by their lower sense of entitlement to higher pay). Interestingly, this finding holds even when the amount that women pay themselves does not come out of what another person receives, thus refuting women's allegedly greater inclination toward caretaking of others as a possible explanation. Id. at 315. These research findings have also been extended beyond lab conditions and generalize to women's pay expectations in the real world. Id. at 317. 70 See Jost, Negative Illusions, supra note 28, at 404-05. 2007] Columbia Journal of Gender and Law far the strongest predictor of gender differences in pay expectations., 71 The tendency to engage in within-gender comparisons also shapes gendered expectations about the distribution of household responsibilities. Working women generally accept unequal household and caretaking labor as legitimate, even while acknowledging that they perform far more such work than their male partners.72 Social psychologists view women's general contentment with their objectively worse job status, pay levels, and household responsibilities as "paradoxical," a product of gender differences in feelings of entitlement.73 The default baseline of drawing comparisons within one's social group has a normalizing effect that downplays or obscures the ways in which our lives are influenced by gender. In her recent essay, The Shadow of Professor Kingsfield. Contemporary Dilemmas Facing Women Law Professors, law professor Martha Chamallas offers one example of how the tendency toward gender-specific comparisons might suppress consciousness of gender bias on a law faculty.7 4 She describes how women on a law faculty turn to their friends, who are likely to be other women, in comparing raises and salaries, which are unpublished at most law schools. She contrasts the normalizing influence of such comparisons with her own observations of large salary disparities on law faculties where some of the highest salaries go to male faculty members who have only average performance records7 5 Gender-specific comparisons also might operate to facilitate bias in the legal academy in other ways. In faculty hiring meetings, women candidates are often implicitly grouped together for comparison with other women candidates, competing for an unarticulated "woman's slot" among the openings for new faculty and implying an incommensurability between 71 See Major, From Social Inequality, supra note 48, at 321. 72 See id. at 296 (explaining that gender, more than any other factor, explains women's disproportionate share of household labor, and yet women in general are not more dissatisfied than men in their relationships and do not perceive their labor overload as unfair). See also KATHARINE T. BARTLETT & DEBORAH L. RHODE, GENDER AND LAW: THEORY, DOCTRINE, COMMENTARY 236-37 (4th ed. 2006) (citing research finding that women do substantially more housework and family caretaking than men and research finding that only a minority of women (one quarter to one third) see this as unfair). 73 See Major, From Social Inequality, supra note 48, at 294-96. 74 Chamallas, The Shadow of Professor Kingsfield, supra note 8, at 203. "5 ld. at 204. [Vol. 16:3696 Perceiving Subtle Sexism male and female candidates.76 Similarly, Professor Derrick Bell has observed an unspoken ceiling on law faculties for African-American hires, in which schools effectively section off "diversity slots," and foreclose serious consideration of African-American candidates once a token level of minority representation has been achieved. 77 Although within-group comparisons might be defended as a necessary by-product of attention to race and gender in the process of promoting diversity and guarding against institutional racism and sexism, such comparisons often operate to reinforce separate and suppressed expectations of excellence for women and people of color and effectively limit the number of slots for which women and minority candidates actually compete. While troubling, I do not believe that the answer to this "double- bind" is to substitute the pretense of gender- and race-neutrality for 78consideration of race and gender diversity in the hiring process. 78 The very intractability and invisibility of race and gender bias in the academy has required aggressive attention to diversity just to give women and people of color a fair chance at equal, not special, consideration.79 Rather, we should increase our vigilance in spotting subtle forms of bias to make sure that women and people of color are not marginalized either by the pretense of race- and gender-blind hiring or as an unintended consequence of attending to race and gender in the process of seeking greater diversity.8° By suppressing expectations and legitimizing existing arrangements, the tendency to make within-group comparisons, like the "just world" ideology and its corollaries, has an inhibitory effect on perceiving and challenging gender bias. 76 For example, I recall instances where colleagues expressed judgments about female candidates with the opening, "among the women, I like [candidate X]." 77 See Derrick A. Bell, Jr., Application of the "Tipping Point" Principle to Law Faculty Hiring Policies, 10 NOVA L. J. 319 (1986). 78 For an interesting discussion of the ubiquitous "double-bind" in feminist theory more generally, see Margaret Jane Radin, The Pragmatist and the Feminist, 63 S. CAL. L. REV. 1699, 1699-1704 (1990). 79 See Merritt & Reskin, supra note 1, at 490-91 (discussing their research showing that aggressive affirmative action "was needed just to assure that faculties identified and hired women who were equal to the white men they so readily hired"). 80 Cf Ruth Anna Putnam, Justice in Context, 63 S. CAL. L. REV. 1797, 1807 (1990) ("Knowledge of obstacles is not itself an obstacle unless it leads to defeatism; for pragmatists it serves as a spur to seek a way to overcome those obstacles."). 6972007] Columbia Journal of Gender and Law C. Cognitive Barriers to Perceiving Individual Instances of Bias on a Case-by-Case Basis The limits of information processing further obstruct the perception of gender bias on an individual level. Cognitive limitations impair our ability to process information in many respects. Social psychologists who study the perception of discrimination have found that people have great difficulty recognizing individual instances of discrimination under normal conditions of information gathering.8 One difficulty stems from people's reluctance to infer discrimination in an individual case where multiple criteria, which allow for multiple interpretations, are used . Faye Crosby, a social psychologist who studies perceptions of discrimination, offers an example to illustrate this problem. When examining a pair of male and female comparators who receive widely varying salaries, and knowing their ratings with respect to the four factors used to determine employee salary (job level, seniority, education, and motivational rating), slight variations among the comparators on any of the four criteria make it impossible to tell for certain whether the pay gap is attributable to gender. Only when presented with a more extensive group picture, showing an overall pay gap between numerous male and female employees, are observers likely to perceive sex discrimination, given that none of the male/female comparators perfectly match up on the relevant criteria.83 Aggregate information is thus needed for the observer to discern inconsistencies in possible sex-neutral explanations and to react skeptically to hypothesized justifications. In real life, however, people rarely have access to the kind of aggregate information that Crosby describes as necessary to enable the recognition of sex discrimination in individual cases. Usually, people know solely of their own personal situations, and perhaps a few anecdotal others. Rarely, and usually only through discovery after filing a lawsuit, do people have access to the kind of organization-wide comparisons that are most helpful in enabling the perception of discrimination. Perhaps more notable, it is not just what is known that matters, but how the information is presented that determines the likelihood of 81 See Crosby, The Denial of Personal Discrimination, supra note 28, at 377; Major et al., Prejudice and Self-Esteem, supra note 31, at 81. 82 See Crosby, The Denial of Personal Discrimination, supra note 28, at 377. 83 See id. at 378. See also Major, From Social Inequality, supra note 48, at 332 ("It is easier to see discrimination on the collective level than on an individual level."). 698 [Vol. 16:3 Perceiving Subtle Sexism perceiving bias. People are more likely to perceive instances of discrimination when they are shown information in the aggregate, all at once, in an across-the-board comparison of the treatment of women and men, and less likely to do so when the same information trickles in piecemeal, presented on an individual case-by-case basis. 84 In other words, formatting is critical. A leading study of this phenomenon used male college students who read case studies of a company and were asked to determine whether the company discriminated against women.8 5 The subjects who read the materials that were formatted on a case-by-case basis perceived less sex discrimination than those who read the same materials presented in an all-at-once, whole-picture format.8 6 Crosby explains that when information is presented in piecemeal fashion, the reviewer hypothesizes rational, nondiscriminatory reasons that might explain each individual case. But when the same information is shown in the aggregate and all at once, the viewer finds hypothetical neutral explanations less credible. 87 This information-processing bias operates at the cognitive, subconscious level, regardless of a person's conscious views about gender.88 Unfortunately, life more often replicates the conditions under which the cognitive limits of information processing are likely to obscure the recognition of gender discrimination. Rarely does evidence of aggregate, across-the-board discrimination present itself all at once, in a neatly packaged format. Rather, like the conditions unfavorable for recognizing discrimination, evidence commonly dribbles in piecemeal over time, usually in the form of individual stories or anecdotes. D. The Reluctance to Challenge Bias and the Effect on Perception The social psychology literature on perceiving and claiming discrimination usually treats these processes as separate and distinct. However, in important respects, the processes that complicate perception 84 See Major et al., Prejudice and Self-Esteem, supra note 31, at 81 (2003). 85 See Faye Crosby et al., Cognitive Biases in the Perception of Discrimination: The Importance of Format, 14 SEx ROLES 637 (1986). 86 Id. at 644. 87 Id. at 645. 88 As Crosby explains, "[t]he cognitive biases that we have demonstrated operate among the 'nice guys' as surely as among the villains." Id. at 646. 2007] 699 Columbia Journal of Gender and Law intersect with and are reinforced by the difficulties that encumber the decision of whether to complain of discrimination. Contrary to the prevailing myth that a person who actually suffered discrimination would have complained, in reality, most people exhibit a deep reluctance to challenge discrimination." 9 Nevertheless, the myth is strong and shapes public reaction to discrimination accounts. It was displayed in full force during the Anita Hill/Clarence Thomas hearings, which spent an inordinate amount of time and energy dissecting Professor Hill's failure to complain about the alleged harassment, and determining what bearing that failure had on her credibility as a witness and on the reliability of her account. 90 The reality is that targets of discrimination rarely complain of discrimination, and for good reason. The costs of complaining are enormous, both in terms of the social costs and the consequences for current and future career opportunities. 91 While lost job opportunities and the fear of retaliation are generally the most frequently cited reasons for not complaining, the role of social costs is important as well.92 Concerns about self-presentation and the reactions of others significantly affect the decision of whether to challenge discrimination.93 People have a general desire to be liked and to appear competent in their professional lives, but people who complain of discrimination are generally neither well-liked nor perceived as 89 See generally Deborah L. Brake, Retaliation, 90 MINN. L. REV. 18, 28-29 (2005) (summarizing and citing evidence documenting the low likelihood of reporting or challenging sexual harassment or sexist behavior). 90 See, e.g., Martha R. Mahoney, Exit: Power and the Idea of Leaving in Love, Work, and the Confirmation Hearings, 65 S. CAL. L. REV. 1283 (1992) (discussing how Professor Hill's failure to leave her job or file a complaint was used against her in the hearings); Louise F. Fitzgerald, Science v. Myth: The Failure of Reason in the Clarence Thomas Hearings, 65 S. CAL. L. REV. 1399, 1402-03 (1992) (explaining how Professor Hill's failure to report the harassment was erroneously used to undermine her credibility). 91 Brake, supra note 89, at 32-36 (describing and citing extensive research documenting the high costs of complaining about discrimination). 92 See id. at 36-42 (discussing the social costs). See also Pamela Hewitt Loy, The Extent and Effects of the Sexual Harassment of Working Women, 17 Soc. Focus 31, 40 (1984) [hereinafter Loy, The Extent and Effects] (describing the results of a study of sexually harassed women finding that the most common negative response to vocalizing harassment was social stigmatization from coworkers, including ignoring and not supporting the women, and that these reactions were very harmful in affecting the work environment and in further stigmatizing the target). 93 Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 168-71. [Vol. 16:3 Perceiving Subtle Sexism competent. 94 Instead, they are often disliked, viewed as troublemakers, or seen as having problematic personalities. 95 Objective evidence that the claimant did, in fact, experience discrimination does not protect people from being disliked when they complain of discrimination. In studies examining predominantly white subjects' reactions to an account in which an African American student blamed poor test results either on discrimination or some nondiscriminatory cause, the subjects disliked the students who blamed discrimination, even when presented with objective evidence that the student had, in fact, experienced racial discrimination in the grading of the test.96 These findings were replicated in follow-up studies in which subjects read accounts of an African American who failed to receive a job offer after an interview with a clearly prejudiced interviewer. 97 Again, the subjects disliked the African Americans who blamed discrimination for the failure to receive a job offer, even when it was clear that the interviewer was prejudiced. Social psychology research has documented a similar disapproval of women who complain of sexism. One notable study found that male subjects liked a woman less when she challenged as sexist a male colleague's remarks that really were sexist, but did not dislike her if she challenged as sexist other remarks that had no apparent connection to gender.98 That is, challenging sexism when it is not really there does not provoke the same degree of hostility, perhaps because such challenges are easily dismissed as non-threatening compared to challenges to sexism that 94 Id. 95 See Brake, supra note 89, at 32. See also White v. Burlington N. & Santa Fe R.R., 364 F.3d 789, 794 (6th Cir. 2004) (upholding a retaliation claim and noting that the plaintiffs employer identified her as a "troublemaker" for complaining of gender discrimination), aff'd, 126 S. Ct. 2405 (2006). 96 See Cheryl R. Kaiser & Carol T. Miller, Stop Complaining! The Social Costs of Making Attributions to Discrimination, 27 PERS. & SOC. PSYCHOL. BULL. 254, 261-62 (2001). 97 See Cheryl R. Kaiser & Carol T. Miller, Derogating the Victim: The Interpersonal Consequences of Blaming Events on Discrimination, 6 GROUP PROCESSES & INTERGROUP REL. 227 (2003). 98 See Dodd et al., Respected or Rejected, supra note 26, at 574-75. See also Alexander M. Czopp & Margo J. Monteith, Confronting Prejudice (Literally): Reactions to Confrontations of Racial and Gender Bias, 29 PERS. & SOC. PSYCHOL. BULL. 532, 541 (2003) (describing the results of a study finding that women's confrontations of sexism generally provoked feelings of hostility or amusement rather than guilt or remorse). 2007] 702 Columbia Journal of Gender and Law [Vol. 16:3 clearly have merit.99 This and other research demonstrates that the fear that claiming discrimination will trigger social and occupational penalties is well-grounded in reality. 00 Despite this reality, people persist in believing that a true victim of discrimination would forcefully complain, and that they themselves would do so if faced with such a situation. One of the more enlightening studies debunking this myth used two groups of female subjects to examine women's beliefs about how they would respond if they experienced discrimination, and then contrasted these beliefs with women's actual responses to discrimination. 01 In this study, one group of women read an account of a job interview in which a male interviewer asked a female interviewee several sexually harassing and discriminatory questions. 10 2 The women were then asked how they would have responded had they been subjected to such questions. The women overwhelmingly predicted that they would have responded assertively, by refusing to answer the questions, confronting the interviewer, or reporting him to a higher authority.10 3 The second part of the study used a different group of female subjects, this time placing them in an actual interview with a male interviewer in which they believed they were being evaluated for a research assistant position. 10 4 These women were then subjected to the same discriminatory questions featured in the earlier part of the study. When 99 See Dodd et al., Respected or Rejected, supra note 26, at 575 ("[P]erhaps [it was] because her unprovoked remark was dismissed as without legitimate cause. On the other hand, the target woman's response to the clearly sexist remark was legitimate, and as a consequence perhaps made especially salient the fact that she was transgressing her gender role by standing up to a man in that situation."). 100 See, e.g., Loy, The Extent and Effects, supra note 92, at 42 (describing the results of a study showing that ignoring sexual harassment was a safer strategy in terms of avoiding negative outcomes than verbalizing the problem to others, including going to a supervisor, addressing a sexual harassment committee, or taking legal action). 101 See Julie A. Woodzicka & Marianne LaFrance, Real Versus Imagined Gender Harassment, 57 J. Soc. ISSUES 15 (2001). 102 Id. at 20-21. 103 A strong majority of the women in this part-of the study (sixty-eight percent) predicted they would refuse to answer at least one of the three sexist questions asked by the interviewer, with sixty-two percent saying they would ask the interviewer why he asked the question or tell him the question was inappropriate, and with twenty-eight percent saying they would take more drastic measures such as immediately leaving the room or confronting the interviewer. Id. at 21. "04 Id. at 21-23. 2007] Perceiving Subtle Sexism 703 actually subjected to the sexually harassing and discriminatory questions, not one of the subjects confronted the interviewer or refused to answer the question.10 5 Instead, the most common response was to simply answer the question.' °6 Some interviewees tried to deflect the question or politely asked why the interviewer was asking a particular question. °107 However, even these interviewees ultimately answered the questions without further objection. 0 8 Follow-up debriefing, confirmed by observations of the interviews themselves, revealed that the most common emotional response at the time was fear, in contrast to the anger that the women in the first part of the study predicted they would feel.'0 9 The study illustrates the stark contrast between how people believe they would respond to discrimination and how they actually respond when they experience it. 110 The gap between expectation and reality creates a dissonance for people who experience prejudice but simultaneously feel a powerful pull not to confront it. In this process, the pressure not to confront bias can have a suppressive effect on the perception of bias. Unless a person is prepared to challenge any bias she perceives, the perception of bias sits uncomfortably with the person's expectation that she would react swiftly and sharply to confront the perpetrator if she experienced discrimination. Minimizing the incident or denying that the experience was in fact discriminatory thus enables the target to excuse or justify her inaction."' By encouraging the 105 More than half of the women ignored the sexist nature of the remark and answered the question, and slightly over one-third politely asked why he had asked the question, but answered it nonetheless. Id. at 23-24. Of this latter group, eighty percent waited until after the interview was over to ask why he had asked the question, and simply responded to it at the moment it was asked. None of the women refused to answer a question, confronted the interviewer, or left the room. Id. 106 id. 107 Id. at 24. 108 Id. 109 Id. at 25. 110 Other studies have produced similar findings. See, e.g., Janet K. Swim & Lauri L. Hyers, Excuse Me-What Did You Just Say?!: Women 's Public and Private Responses to Sexist Remarks, 35 J. EXPERIMENTAL SOC. PSYCHOL. 68 (1999) (comparing women's predicted confrontational responses to a male colleague's sexist remarks with their actual silence in the face of such behavior). '11 Quinn, The Paradox of Complaining, supra note 27, at 1171 (describing an "internal dialogue gauging whether one should be and can afford to be aggrieved by the action" and explaining how and why sexually harassed women often choose "a story that explains away one's own impotence"). Columbia Journal of Gender and Law target to revise and recheck her perceptions to ease the clash between expected and actual behavior, the failure to confront may subtly undermine the perception of bias. 1 2 The hefty pressures not to confront discrimination operate to submerge consciousness of gender bias at the collective level as well. Social psychologist John T. Jost discusses the problem of "pluralistic ignorance," which occurs when individuals suppress or deny their awareness of a problem in part because they interpret the silence of others to confirm its absence." 3 When people are reluctant to publicly identify or challenge gender bias within institutions, their silence contributes to a normalcy in which individuals interpret their own experiences and perceptions consistently with that collective silence. In this way, the silence of others operates to inhibit and potentially undermine an individual's perception of bias.' 1 4 E. A Complication: The Relationship Between Self-Esteem and Perceiving Bias The processes described above all operate in the direction of submerging awareness of gender bias. However, one motivational incentive that surfaces in the literature works in the other direction, possibly encouraging attributions of bias. Some researchers have hypothesized that attributions to discrimination might serve to protect and preserve self- esteem.1 5 In this account, the drive for self-esteem creates an incentive for people to blame negative outcomes on discrimination in order to avoid the 112 See Olsen & Hafer, Tolerance of Personal Deprivation, supra note 28, at 164 (citing theorists who explain the gap between perceptions of discrimination against one's group and oneself because such denial "allows the disadvantaged individuals to justify inaction against the more powerful"). Cf. Major et al., Prejudice and Self-Esteem, supra note 31, at 87 ("[O]ne method of coping is cognitive reappraisal-redefining an event as less threatening than it was originally appraised as being."). 113 Jost, Negative Illusions, supra note 28, at 406-07. 114 For one example from a reported decision of how collective silence is used to undermine an individual's perception of bias, see Davis v. Monroe County Bd of Educ., 526 U.S. 629, 635 (1999) (reciting an alleged fact in a sexual harassment case that the school principal responded to a student's allegation of sexual harassment by asking why she "was the only one complaining"). Cf Quinn, The Paradox of Complaining, supra note 27, at 1155 ("Like so many gendered and sexually related harms, if you don't talk about it, it hasn't happened."). 115 See Kaiser & Major, supra note 25, at 803 (summarizing research); Major et al., Attributions to Discrimination, supra note 9, at 220-22 (surveying literature). [Vol. 16:3 Perceiving Subtle Sexism blow to self-esteem that would occur if they located the blame internally. This theory is worth some discussion because of its potential to offset the above processes and support the alternative hyper-vigilance thesis. The more social psychologists study the connection between attributions to discrimination and the effect on self-esteem, the more complicated that relationship appears. 1 6 There is much debate within the literature about how conceptions of self-esteem enter into and interact with the process of perceiving discrimination. Some literature does support the view that blaming negative outcomes on discrimination has a positive effect on self-esteem. 1 7 Under this view, individuals might favor attributions to discrimination as a way of protecting self-esteem from the damage that alternative, internally-directed explanations might cause. However, recent work has cast doubt on such a simplistic understanding of how self-esteem interacts with perceptions of discrimination. Some researchers have attempted to more finely parse self- esteem, separating one's sense of self with respect to individual achievement from one's sense of self as a social actor. 18 While attributions to discrimination may preserve positive feelings about one's abilities, they may harm a more socially oriented self-esteem by undermining one's sense of self as a social actor."19 This negative effect on social self-esteem is especially pronounced for individuals who strongly identify with their social group.120 This complication may call into question the causal 116 See Major et al., Attributions to Discrimination, supra note 9, at 222-23 (summarizing existing literature); Major et al., Prejudice and Self-Esteem, supra note 31 (positing a transactional model in which numerous situational factors mediate the relationship between perceiving discrimination and self-esteem); Brenda Major et al., It's Not My Fault: When and Why Attributions to Prejudice Protect Self-Esteem, 29 PERS. & SOC. PSYCHOL. BULL. 772, 779-80 (2003) [hereinafter Major et al., It's Not My Fault]. 117 See, e.g., Major et al., It's Not My Fault, supra note 116 (citing studies showing a positive correlation between perceiving discrimination and protecting self-esteem, and explaining their finding that attributing negative outcomes to discrimination promotes self- esteem by locating the "fault" with an external cause rather than an internal one; e.g., the professor excluded me because of sexism vs. the professor excluded me because he thinks I am stupid). Interestingly, however, even when attributions to discrimination shelter self- esteem, they do not protect against feelings of hostility or anxiety. Id. 118 See Major et al., Attributions to Discrimination, supra note 9, at 222-23; Major et al., Prejudice and Self-Esteem, supra note 31, at 78, 84-85. "19 Major et al., Attributions to Discrimination, supra note 9, at 222 (citing research showing that perceiving bias hurts self-esteem since one's social group membership is an important part of the self). 120 Major et al., Prejudice and Self-Esteem, supra note 3 1, at 93. 2007] 705 Columbia Journal of Gender and Law relationship between preserving self-esteem and perceiving discrimination altogether. As explained later in this Article,' 2 1 strongly identifying with one's social group increases the likelihood of perceiving bias on the basis of membership in that group. Yet, having a strong identification with one's social group makes it more likely that perceiving oneself as a target of bias will hurt self-esteem.12 Thus, strong social group identification correlates both with an increased likelihood of perceiving bias and with an increased likelihood that perceiving bias will harm self-esteem. If the desire to protect self-esteem strongly motivated people to perceive discrimination, one would expect people who strongly identify with their social group to be less likely to perceive discrimination, since for this group, perceiving bias hurts a core aspect of their self-esteem. The influence of social group identification on perceiving bias, and its role in sorting out the self-esteem consequences for persons who do so, suggests reason to doubt that the goal of protecting self-esteem strongly encourages the perception of discrimination. The most recent work in this area suggests a complicated relationship between self-esteem and perceiving bias, one largely dependent on the extent of the discrimination and the clarity and openness of prejudice cues.12 3 Self-esteem does not benefit when people perceive that they have experienced systematic and pervasive discrimination. 24 Any positive effect on self-esteem from picking prejudice over an alternative, internal explanation is substantially limited to where the perceived prejudice involves a discrete and specific event. 25 The subtlety of prejudice cues also affects the relationship between self-esteem and blaming discrimination. People who perceive sexism when exposed to obvious signs of prejudice show higher self-esteem than those who perceive prejudice when exposed to more subtle signs of prejudice.' 2 6 121 See infra notes 133-135 and accompanying text. 122 See Major et al., Prejudice and Self-Esteem, supra note 31, at 93-94. 123 Major et al., Attributions to Discrimination, supra note 9, at 221 (explaining that the complex relationship between perceiving discrimination and protecting self-esteem is best explained by situational factors, especially how blatant the prejudice was and whether the circumstances were ambiguous). 124 Id. at 230. 125 Id. at 229; Major et al., Prejudice and Self-Esteem, supra note 31, at 90-91. 126 Major et al., Attributions to Discrimination, supra note 9, at 229. In this study, the condition for ambiguous prejudice cues involved a creativity test where a female cohort says to the subject: "You know, I have friends who were in this study, and they told me that 706 [Vol. 16:3 2007] Perceiving Subtle Sexism 707 When prejudice cues are subtle or ambiguous, perceiving sexism does not exclude other explanations that simultaneously locate fault internally.' 27 Under ambiguous conditions, people tend to infer multiple causes for negative outcomes, finding prejudice and internal shortcomings jointly responsible.1 28 Thus, when uncertainty and ambiguity are present, self- esteem is not protected by perceiving prejudice. 129 The bottom line is that people who blame discrimination for the bad outcomes in their lives do not necessarily receive a self-esteem boost. The connection between perceiving bias and protecting self-esteem is complicated, to say the least, and it does not appear strong enough to rebut the under-perception thesis or offset the effects of the processes described above which operate to suppress perceptions of bias. 130 the guy doing the evaluating totally grades guys and girls differently." Id. at 224. In the overt prejudice group, the female cohort says: "You know, I have friends who were in this study, and they told me that the guy doing the evaluating is totally prejudiced. He never picks a girl to be the team leader-he always picks a guy." Id. Under both conditions (and a third condition with no prejudice cue), the male evaluator assigns low grades to the female subjects, who are then evaluated for how they explain their poor results and the effect on their self-esteem. Id. at 224-25. See also Major et al., Prejudice and Self-Esteem, supra note 31, at 91 (discussing the relationship between the intensity of prejudice cues and the effect on self-esteem from perceiving prejudice). 127 Major et al., It's Not My Fault, supra note 116, at 775 (noting that "attributions to discrimination and self-blame are not necessarily inversely related" and "[p]erceiving that another person is prejudiced against one's group does not preclude blaming a negative outcome on aspects of oneself, such as one's lack of effort"). 128 Major et al., Attributions to Discrimination, supra note 9, at 229. 129 Id. As the authors explain: Uncertainty is an affectively unpleasant state, and may lead to reduced well-being relative to being confident that one's outcomes are due to prejudice. When one is faced with blatant prejudice, in contrast, there is no uncertainty about the cause of one's outcomes. It is clear that the proper attribution for negative outcomes is prejudice rather than one's lack of ability. Id. at 229 (citations omitted). 130 Id. at 230 ("This study adds to growing evidence that it is overly simplistic to assert that attributing negative outcomes to prejudice protects self-esteem .... Ironically, the self-protection component of making attributions to discrimination applies most clearly when the situation is unambiguously prejudiced .... In the real world, ambiguity is likely to be the rule, rather than the exception."). Columbia Journal of Gender and Law F. The Limits of Generalizing: Why Do People Ever Perceive Discrimination? In considering the psychological forces that constrain the ability of individuals to perceive subtle forms of bias, it is important to keep in mind that these processes are not a monolithic set of constraints that have the same effect on all persons, nor are they insurmountable. Notwithstanding the processes described above that create "static" or "interference," some people do perceive that they have experienced gender or race bias. This reality raises an important question: why are some people more likely than others to interpret an event as discriminatory? Existing research identifies two factors that increase the likelihood that an individual will perceive bias: a skepticism of "just world" ideology and a high level of group-identification with members of one's social group. With respect to the first factor, not everyone shares a strong belief in the existence of a "just world."' 131 Although more research is needed to determine what influences individuals' adherence to "just world" ideology, some research has found that different social groups respond to "just world" ideology differently. For example, African Americans as a group are less likely than women as a group to believe in "just world" ideology. 132 However, this finding highlights a larger problem with the research in this area. The focus on monolithic social groups misses important lessons that have emerged from the critique of intersectionality by Critical Race Feminists. By examining women and African Americans as discrete groups, the research is unable to account for how race and gender together may shape adherence to "just world" ideology and influence perceptions of bias. More sophisticated attention to the intersection of race and gender by researchers might help further our understanding of how and why some people come to believe more ardently in "just world" ideology than others. 133 The second factor, an individual's level of social group identification and feelings of solidarity, also emerges as an important influence on the likelihood of perceiving discrimination based on social 131 Major, From Social Inequality, supra note 48, at 335. 132 See id at 336-37. 133 Cf Tanya Kateri Hernndez, The Intersectionality of Lived Experience and Anti-Discrimination Empirical Research, in HANDBOOK OF EMPLOYMENT DISCRIMINATION RESEARCH, supra note 26, 325, 325-35 (criticizing the lack of attention to the intersection of race and gender in social science research on sexual harassment and calling for greater attention to intersectionality in social science research on discrimination). [Vol. 16:3 Perceiving Subtle Sexism group membership. Strongly identifying with one's social group broadens one's perspective beyond the individual level and encourages comparisons with out-group members, thereby facilitating perceptions of bias. 3 4 In one study, the extent to which women identified with women as a social group significantly affected their likelihood of perceiving discrimination under conditions where prejudice cues were ambiguous. 35 Women who had declared in an earlier part of the study that their gender was a significant element in their social identity were significantly more likely to perceive that they had been subjected to subtly sexist treatment than women who ranked gender lower as an element of their identity. This finding is hardly surprising. One would expect, for example, women who call themselves feminists and strongly identify with other women to be more inclined to see gender bias in daily life. Notably, however, this research does not lend support to those critics who blame feminism's emphasis on gender identity for turning everything into a gender issue. The study's most interesting finding was that having a strong identification with women as a social group had no effect on attributions of bias when prejudice cues were either blatant or absent. 136 Thus, an affinity with one's social group plays a strong role in influencing perceptions of subtle bias, but does not induce perceptions of bias where there is no evidence of prejudice. Although the findings on social group solidarity are helpful in developing an understanding of why some people infer bias from a situation while others in the same social group do not, this research too is limited by its failure to more finely parse social group identity formation. By failing to examine how an individual's gender identity intersects with her racial identity or other social identities important to her sense of self, the research 134 See Major, From Social Inequality, supra note 48, at 331 ("Shifting people's perspective from their personal situation to that of their group enhances feelings of discrimination."). 135 See Major et al., Attributions to Discrimination, supra note 9, at 225-26, 228- 29. See also Major et al., Prejudice and Self-Esteem, supra note 31, at 95 (discussing research showing a positive association between having a high level of group identification and the likelihood of perceiving prejudice); id. at 94 (explaining that group-identification can help individuals cope with perceived discrimination since "[g]roups can provide emotional, informational and instrumental support, validation for one's perceptions, and social consensus for one's attributions"). 136 Major et al., Attributions to Discrimination, supra note 9, at 228 ("Increased identification with one's group may shift interpretation of ambiguous events from the individual to the group level ... and make individuals more aware of the potential for group injustice."). 2007] Columbia Journal of Gender and Law assumes that an individual's gender consciousness develops and operates independent of her race. All women belong to multiple social groups, and we should not assume that a woman's identification with women as a social group can be understood in unidimensional terms, or that membership in other social groups would not simultaneously influence different formulations of group solidarity. A person's identification with a social group is far more complex than can be understood by examining a single layer of group identity in isolation. More research is needed to explore how both of these factors- adherence to "just world" ideology and allegiance to one's social group- might be marshaled to increase the accuracy of perceptions of bias. What forces encourage group-identification and solidarity? Perhaps counterintuitively, one study suggests that increasing consciousness of the prevalence of discrimination against one's social group does not necessarily promote a strong identification with one's social group.137 In this study, Latino and Latina students who had previously reported a low degree of group identification with Latino Americans later identified even less with their social group after reading about pervasive discrimination against Latino Americans. 138 The reverse was true for Latino and Latina students who had previously identified strongly as Latino Americans; they identified with Latino Americans even more strongly after reading this account. 139 These findings suggest that an individual's development of group- consciousness is a complicated process. Upon reflection, the fact that some of the students in the study would react to the accounts of widespread discrimination by distancing from their social group is neither surprising nor irrational. As a strategy, it seems designed to preserve an individual's sense of agency and control by discounting the significance of membership in a stigmatized group to that individual's identity. Such distancing occurs in many settings, including in the legal profession. For example, after the release of a study identifying gender bias in some segment of the legal profession, 40 some women may react negatively to distance themselves 137 See Major et al., Prejudice and Self-Esteem, supra note 31, at 95. 138 id. 139 id. 140 See, e.g., PHYLLIS G. KITZEROW & VIRGINIA M. TOMLINSON, THE LEGAL PROFESSION: A STUDY OF ACBA MEMBERSHIP 2005 (2006), available at http://www.acba.org/ACBA/pdf/ACBAMemSurveyResults2005.pdf (survey of Allegheny County Bar Association members showing persistent disparities in pay and job satisfaction between male and female attorneys). For public reaction to the survey, see also Gary 710 [Vol. 16:3 Perceiving Subtle Sexism from the results of the study, questioning its methodology and proclaiming that they themselves have never experienced gender bias in the profession. Such distancing can make it difficult to talk about perceptions of bias and contributes to a silence that normalizes the status quo. For example, a senior female colleague might respond to a more junior colleague's story of her experiences with male students or colleagues dismissively, with a tone that suggests "I would never let that happen to me," and in doing so, send a message denying any influence of gender and individualizing the problem. Such a response, by distancing from any common gender experience, individualizes the problem and makes it difficult to discuss or assess how gender may have influenced the experience. A similar process might be set in motion by a senior person of color on a law faculty who heaps criticism on an African American candidate, opening the door to and legitimating a tirade of subsequent criticism, with the effect of preempting questions about whether the candidate is being fairly considered or concerns about institutional bias. Yet, many individuals do not engage in such distancing strategies and instead embrace a commonality of interests and a sense of collective destiny with members of their social group. Questions about how group consciousness develops and its role in facilitating perceptions of bias, and constructive responses to it, deserve further study. II. DISCRIMINATION LAW AND ITS INFLUENCE ON THE PERCEPTION OF GENDER BIAS Much legal scholarship falling under the wide umbrella of the "law and society" movement asserts and explains the thesis that law actively participates in the production of knowledge.14 1 Legal narratives and Rotstein, Women Lawyers Here Still Lagging in Pay, PITTSBURGH POST-GAZETTE, Sept. 14, 2006, http://www.post-gazette.com/pg/06257/721762-28.stm. 141 See, e.g., Austin Sarat & Thomas R. Kearns, Beyond the Great Divide: Forms of Legal Scholarship and Everyday Life, in LAW IN EVERYDAY LIFE 21, 27- 28, 50-54 (Austin Sarat & Thomas R. Kearns eds., 1993) (discussing law's "constitutive force" and "meaning- making power" as explored in legal scholarship); CAROL SMART, FEMINISM AND THE POWER OF LAW 10-13 (1989) (developing a feminist critique of legal knowledge and explaining how legal discourse stakes out a special claim to truth and disqualifies other forms of knowledge and experience); MARIANA VALVERDE, LAW'S DREAM OF A COMMON KNOWLEDGE 5-15 (2003) (explaining her approach, which draws from sociology of knowledge studies, in exploring law's production of knowledge); Lawrence Douglas et al., Theoretical Perspectives on Lives in the Law, in LIVES IN THE LAW 1, 10-11 (Lawrence Douglas et al. eds., 2002) (describing the work of socio-legal scholars emphasizing law's role in the production of knowledge). 2007] Columbia Journal of Gender and Law discourses do more than passively reflect social and extralegal knowledge; they actively shape and influence it. To the extent that discrimination doctrine and the narratives that emerge from it shape our understandings and perceptions of bias, their messages work to conceal rather than expose the existence of gender bias. Specifically, discrimination law's narratives reinforce the cognitive and motivational barriers to perceiving bias described in the previous section. The dominant story told by discrimination law is one of a world in which merit is the rule and discrimination is aberrational and insignificant as an explanation for pervasive inequality in outcomes. 42 The universe of actionable discrimination is, for the most part, limited to conscious, intentional discrimination by an individual perpetrator. 143 It is eclipsed by the alternative and much larger universe of merit and individual responsibility. Discrimination cases are framed as stories with competing, mutually exclusive narratives in which the fact-finder has to decide between the story line in which a bad, intentionally sexist or racist actor harms an innocent victim and the one in which the plaintiff's own flaws produced the bad ending. These story lines clash in countless discrimination cases. More often than not, the narrative that confirms the "just world" and individual responsibility thesis prevails. 44 In the dominant narrative, the plaintiff failed to get the raise, promotion, or opportunity because she was not sufficiently competent or deserving. 45 On the fewer occasions when the 142 See, e.g., Kimberl6 Williams Crenshaw, Race, Reform and Retrenchment: Transformation and Legitimation in Antidiscrimination Law, 101 HARV. L. REV. 1331 (1988); Alan Freeman, Antidiscrimination Law: The View from 1989, 64 TUL. L. REV. 1407 (1990); Mary E. Becker, Obscuring the Struggle: Sex Discrimination, Social Security, and Stone, Seidman, Sunstein & Tushnet's Constitutional Law, 89 COLUM. L. REV. 264 (1989). 143 See, e.g., Chamallas, Deepening the Legal Understanding, supra note 32, at 747-48; Samuel R. Bagenstos, The Structural Turn and the Limits of Antidiscrimination Law, 94 CAL. L. REV. 1, 4-10 (2006) (summarizing scholarship criticizing antidiscrimination law's focus on intentional discrimination and its failure to reach unconscious bias). 144 See, e.g., Michael Selmi, Why Are Employment Discrimination Cases So Hard to Win?, 61 LA. L. REV. 555, 559-62 (2001) [hereinafter Selmi, Why Are Employment Discrimination Cases So Hard to Win?] (examining lower court decisions in employment discrimination cases and insurance cases and finding that plaintiffs fare much worse in employment discrimination cases); Catherine J. Lanctot, Secrets and Lies: The Need for a Definitive Rule of Law in Pretext Cases, 61 LA. L. REV. 539, 546-49 (2001) (critiquing the anti-plaintiff bias in employment discrimination cases). 145 For a discussion of how discrimination law reinforces this dominant narrative, see West, supra note 41, at 145: 712 [Vol. 16:3 2007] Perceiving Subtle Sexism 713 alternative narrative prevails, there are usually enough smoking gun comments to suggest that one or more bad actors intentionally and malevolently took sex into account. 146 In viewing bias-versus-merit as an either/or system of mutually exclusive story lines, law confines discrimination to the tiny space reserved for anomalous exceptions to the "just world" norm. 147 In this way, discrimination law purports to rectify aberrations from a presumptive meritocracy, while ultimately reinforcing the dominant story line that "people get what they deserve." In a reciprocal relationship, the belief in a "just world" and individual responsibility shapes law's stingy approach to The intent requirement in discrimination law is one way of maintaining and justifying the present myth of meritocracy. By pretending that discrimination results only from decisions made by people with prejudicial motives, and because most of us think we do not operate with such motives, we can continue to believe that those of us who have succeeded in this system have done so by our own meritorious performance. Id. 146 See, e.g., Price Waterhouse v. Hopkins, 490 U.S. 228, 272-73 (1989) (O'Connor, J., concurring) (relying on sexist remarks to infer stereotyping). However, even when overtly sexist remarks occur, courts often minimize the significance of such remarks and reject them as sufficient to establish discrimination. See Anne Lawton, The Meritocracy Myth and the Illusion of Equal Employment Opportunity, 85 MiNN. L. REV. 587, 634-42 (2000) (criticizing a trend in the lower courts to discount overtly biased remarks). 147 Although the mixed motive model of employment discrimination law purports to address a situation where multiple motives, legitimate and illegitimate, caused the adverse outcome, its cramped application in the courts has relegated it to the margins. See Ann C. McGinley, iViva La Evoluci6n!: Recognizing Unconscious Motive in Title VII, 9 CORNELL J.L. & PUB. POL'Y 415, 472-77 (2000) (describing this model and its limitations). Moreover, this model is equally vulnerable to the pull of "just world" ideology, since it assumes that merit and sexism are distinct and separable categories. For example, in Price Waterhouse v. Hopkins, the paradigmatic mixed motive case, Justice Brennan's plurality opinion assumed that the plaintiff's interpersonal problems provided a legitimate basis for the denial of partnership, as long as this reason alone, untainted by gender stereotyping, would have caused it to make the same decisions. 490 U.S. 228, 252 (1989). Justice Brennan stated: The very premise of a mixed motives case is that a legitimate reason was present, and indeed, in this case, Price Waterhouse already has made this showing by convincing Judge Gesell that Hopkins' interpersonal problems were a legitimate concern. The employer instead must show that its legitimate reason, standing alone, would have induced it to make the same decision. 714 Columbia Journal of Gender and Law [Vol. 16:3 discrimination claims, and the narrow confines of discrimination doctrine in turn reinforce this very ideology.148 In the process, the messages of discrimination law contribute to the denial of gender bias by bolstering the ideologies that suppress attributions to bias. The dominant narratives of discrimination law also reinforce the reluctance to detect bias by limiting actionable discrimination, for the most part, to instances where an individual decision-maker acts with a conscious, discriminatory intent.149 Despite calls to broaden the reach of discrimination doctrine, attempts to incorporate a structural or institutional approach into discrimination law have succeeded only at the margins.150 The dominant frame of discrimination law still requires a villain. Under such a governing framework, it is difficult to resist the temptation to set aside suspicions of bias in the face of protestations that the actor who might have engaged in it is "a nice guy.,' ' 5 1 More subtly, discrimination doctrine also reinforces the appropriateness of within-gender comparisons. Although actionable discrimination claims require a plaintiff to establish differential treatment in relation to a similarly situated person of the other sex, discrimination law treats such comparisons as legitimate only when the comparators are similar 148 See, e.g., Selmi, Why Are Employment Discrimination Cases So Hard to Win?, supra note 144 (attributing the unusually high loss record of employment discrimination plaintiffs to judges' unconscious bias and reluctance to believe that discrimination is widespread); Michael Selmi, Proving Intentional Discrimination: The Reality of Supreme Court Rhetoric, 86 GEO. L. REV. 279 (1997) (arguing that the Court's application of a stringent discriminatory intent requirement reflects its skepticism about the continuing frequency and extent of discrimination in society). 149 See Lawton, supra note 146, at 590-91; McGinley, supra note 147, at 445-46, 480 (summarizing social science research on unconscious bias and criticizing Title V1I's limitation to conscious, intentional bias). See also Tristin K. Green, Discrimination in Workplace Dynamics: Toward a Structural Account of Disparate Treatment Theory, 38 HARV. C.R.-C.L. 91, 136-44 (2003) (discussing the limits of disparate impact doctrine). 150 See, e.g., Bagenstos, supra note 143, at 2-4 (citing scholarship in the past decade calling for a structural, regulatory approach to discrimination law, but observing that these efforts have failed to make major inroads in changing the shape of discrimination law). 151 See generally Quinn, The Paradox of Complaining, supra note 27, at 1151 (discussing the strategies women use to deflect and reframe harassment by not "taking it personal"). For an extreme example of one plaintiff's prolonged but ultimately unsuccessful efforts to reframe and excuse sexist behavior with the "nice guy" label, see Lamere v. New York State Office for the Aging, No. 03-CV-356, 2005 U.S. Dist. LEXIS 41904 (N.D.N.Y. Apr. 27, 2005). See also Zelinski v. Pennsylvania State Police, No. 03-4025, 2004 U.S. App. LEXIS 16576 (3d Cir. Aug. 11, 2004) (noting that the supervisor dismissed the plaintiff's harassment allegation because the alleged harasser was a "nice guy"). 2007] Perceiving Subtle Sexism 715 except for their gender. The case law overwhelming questions the salience of comparing women's treatment to that of men where women are differently situated from men. 152 The doctrinal take-home message is that a proper comparison is one that compares likes with likes, often referred to as the similarly situated requirement. Discrimination law also has the potential to suppress women's sense of entitlement by treating so much gender inequality as not actionable, implicitly classifying what remains as sex-neutral. In making this claim, I recognize that law's effect on individual sense of entitlement is complicated, as legal rights can empower people to expect and demand more favorable treatment. 153 However, a narrow construction of rights can suppress feelings of entitlement by normalizing and legitimating the range of conduct that does not fall within the scope of legal rights. A number of legal doctrines within discrimination law operate to narrow the scope of gender equality rights, setting a high bar for what is actionable, and implicitly erasing gender bias that is not encompassed by the reach of the law. Harassment law, for example, reaches only conduct that is severe or pervasive, sheltering even offensive gender-based behavior that falls short of that standard. 154 The requirement of tangible harm for vicarious liability without an affirmative defense also promotes the idea that the worthiness of the claim turns on the level of harm. 155 Absent tangible 152 For two of many examples, see Stanley v. University of Southern California, 178 F.3d 1069 (9th Cir. 1999) (holding that the University's decision to pay a female coach of a women's team less than a male coach of a men's team was not discriminatory where the male coach was differently situated with respect to bringing in revenue), and Ezold v. Wolf Block, Schorr and Solis-Cohen, 983 F.2d 509 (3d Cir. 1992) (overturning a lower court's finding that the law firm discriminated against the plaintiff in not making her partner where the plaintiff was not sufficiently similar to any of the male associates who made partner). See also CATHARINE A. MACKINNON, TOWARD A FEMINIST THEORY OF THE STATE 200, 220-21 (1989) (criticizing the similarly situated requirement in sex discrimination law); Catharine A. MacKinnon, Reflections on Sex Equality Under Law, 100 YALE L. J. 1281 (1991) (same). 153 For an excellent discussion of how the Family and Medical Leave Act has increased employees' bargaining power in informal negotiations with employers, see Catherine R. Albiston, Law in the Workplace: Bargaining in the Shadow of Social Institutions: Competing Discourses and Social Change in Workplace Mobilization of Civil Rights, 39 LAW & Soc'Y REV. 11 (2005). 154 See Theresa M. Beiner, Let the Jury Decide: The Gap Between What Judges and Reasonable People Believe is Sexually Harassing, 75 S. CAL. L. REV. 791, 805-19 (2002) (arguing that courts set the bar too high in determining what level of severity and pervasiveness is required to qualify as unlawful harassment). 155 See Burlington Indus., Inc. v. Ellerth, 524 U.S. 742, 761 (1998); Faragher v. City of Boca Raton, 524 U.S. 775 (1998). 716 Columbia Journal of Gender and Law [Vol. 16:3 harm, harassment victims bear the costs of not complaining and failing to take the initiative to improve their situation. 156 Title VII case law also includes a line of cases that carve out de minimus discrimination from the law's reach, legalizing sex bias that stops short of a certain threshold of harm. 157 Finally, retaliation law, despite the Court's recent rejection of an ultimate decision requirement, still requires a sufficient level of adverse action so as to deter a reasonable person from complaining. The Court's decision in Burlington Northern & Sante Fe Railroad Company v. White, 158 which formulated this standard, is too recent to tell how much of a hurdle it will create in retaliation cases; however, the Court's suggestion that mere social ostracism by coworkers would not likely suffice indicates that some forms of retaliation will fall short of this standard. 159 The underlying message in all of this doctrine reminds me of the best-selling self-help book from years back, Don't Sweat the Small Stuff... and It's All Small Stuff 160 The disconnect between the rhetoric and reality of retaliation law also fuels the reluctance to challenge bias, which in turn encourages people 156 See Pennsylvania State Police v. Suders, 542 U.S. 129, 143 (2004) (explaining the Court's sexual harassment precedents as allowing employers the opportunity to establish an affirmative defense to liability for creating a hostile environment absent tangible harm to the plaintiff). 157 For a summary of cases exempting "minor" adverse employment decisions from challenge, see MICHAEL J. ZIMMER ET AL., CASES AND MATERIALS ON EMPLOYMENT DIScRiMINATION 104-05 (Aspen, 6th ed. 2003) (including as "minor" such actions as asking gender-specific questions at job interviews, giving an employee a mid-range evaluation, laterally transferring an employee without a cut in pay, failure to provide a computer, replacing personal staff and giving a negative evaluation, and the denial of a bonus). See also Rebecca Hanner White, De Minimus Discrimination, 47 EMORY L. J. 1121 (1998) (criticizing the de minimus exception in Title VII law); Theresa M. Beiner, Do Reindeer Games Count as Terms, Conditions or Privileges of Employment Under Title VII?, 37 B.C. L. REV. 643 (1996) (criticizing Title Vii's failure to reach outside-the-workplace networking). "' 126 S. Ct. 2405, 2409 (2006) (concluding that Title VIl's prohibition on retaliation "covers those (and only those) employer actions that would have been materially adverse to a reasonable employee or job applicant," and observing that, "[i]n the present context that means that the employer's actions must be harmful to the point that they could well dissuade a reasonable worker from making or supporting a charge of discrimination"). 159 Id. at 2415 (distinguishing "significant" harms, which are covered, from "trivial harms," which are not, and suggesting that "snubbing" and "antipathy" from coworkers would fall in the latter category). 160 RICHARD CARLSON, DON'T SWEAT THE SMALL STUFF... AND IT'S ALL SMALL STUFF: SIMPLE WAYS TO KEEP THE LITTLE THINGS FROM TAKING OVER YOUR LIFE (1997). Perceiving Subtle Sexism to submerge perceptions of bias as a way of explaining why they did not complain. If sexism strikes and no one complains, perhaps it wasn't sexism at all. As explained in the previous section, a person who perceives discrimination but does not challenge it defies widely shared expectations that true victims of bias confront and challenge it. But in reality, few people challenge bias when it targets them. The discomfort of perceiving bias but not challenging it can set in motion a process of revision and denial. Retaliation law does not offer enough encouragement to prospective complainants to break through this process. Discrimination law is replete with generous and reassuring rhetoric promising full protection from retaliation, which reinforces the pressure on victims to complain and harsh judgments against them when they do not. 161 But the law's reality does not match its rhetoric. Legal protections fall far short of offering full protection to complainants, leaving fears of retaliation well-founded. This Article is not the place for a thorough discussion of the limits of retaliation law, but a number of doctrinal limitations deserve brief mention. As I have detailed elsewhere, the reasonable belief doctrine imposes a stark limit on plaintiffs who complain of conduct they perceive as racist or sexist, but which the current legal doctrine does not recognize as actionable discrimination.1 62 Applying this doctrine, the Fourth Circuit recently ruled that an African American who was fired for complaining of a coworker's blatantly racist comments in a break room had no recourse because it was not reasonable for him to believe that these isolated comments, in and of themselves, created a legally actionable environment. 63 As lower courts have applied the reasonable belief doctrine in retaliation cases, it leaves people woefully unprotected from retaliation, especially when they challenge subtle forms of bias that do not perfectly match narrower judicial understandings of actionable discrimination.' 64 Other limits also leave many persons who complain of discrimination unprotected from retaliation. Stringent approaches to 161 See, e.g., Burlington Northern, 126 S. Ct. at 2414 (stating that Title VII provides "broad protection" from retaliation); Matvia v. Bald Head Island Mgmt., Inc., 259 F.3d 261, 270 (4th Cir. 2001) ("The bringing of a retaliation claim, rather than failing to report harassment, is the proper method for dealing with retaliation.") (internal citation omitted). 162 Brake, supra note 89, at 76-103. 163 See Jordan v. Alternative Resources Corp., 467 F.3d 378 (4th Cir. 2006). 164 See Brake, supra note 89, at 86-101. 2007] 717 Columbia Journal of Gender and Law [Vol. 16:3 causation can make it difficult to infer retaliation without a smoking gun. 165 In addition, not all retaliatory actions are prohibited, only those which courts believe would chill a reasonable person from coming forward with a complaint.166 There is a danger that courts will underestimate the propensity for even seemingly minor retaliatory acts to chill potential complaints, as evidenced by the Court's questionable assumption that "snubbing" or social ostracism by colleagues would not be likely to do SO.167 An additional constraint limits the type of actions challengers may take in confronting discrimination and still receive protection from retaliation. Professor Terry Smith has insightfully critiqued the requirement that opposition to discrimination must take the form of a "reasonably proportionate" response, showing how courts have used this doctrine to sort out some challengers, especially persons of color, as undeserving of protection.' 68 Finally, 165 See, e.g., Stone v. City of Indianapolis Pub. Utils. Div., 281 F.3d 640, 644 (7th Cir. 2002) (requiring employee to show either direct evidence of retaliation or that "only he, and not any similarly situated employee who did not file a charge, was subjected to an adverse employment action even though he was performing his job in a satisfactory manner"). Proof of causation is further complicated by the refusal of many courts to extend § 703(m) of the Civil Rights Act to retaliation claims, thereby allowing employers to defeat liability by proving that, notwithstanding the existence of a retaliatory motive behind the adverse action, the same action would have been taken anyway for a legitimate reason. See, e.g., Pennington v. City of Huntsville, 261 F.3d 1262 (11 th Cir. 2001); Speedy v. Rexnord Corp., 243 F.3d 397 (7th Cir. 2001); Matima v. Celli, 228 F.3d 68 (2d Cir. 2000); Kubicko v. Ogden Logistics Servs., 181 F.3d 544 (4th Cir. 1999); Woodson v. Scott Paper Co., 109 F.3d 913, 933-35 (3d Cir. 1997). 166 Burlington Northern, 126 S. Ct. at 2414-16. 167 Id. at 2415. For evidence suggesting that this assumption is incorrect, see Loy, The Extent and Effects, supra note 92, at 40 (describing the results of a study finding that the most common negative response to the vocalization of sexual harassment by sexually harassed women was social stigmatization from coworkers, including ignoring and nonsupport, and that these reactions negatively affected the work environment and further stigmatized the target). 168 See Terry Smith, Everyday Indignities: Race, Retaliation, and the Promise of Title VII, 34 COLUM. HUM. RTs. L. REV. 529 (2003). For a sampling of cases employing this doctrine to exempt complainants from protection, see Jennings v. Tinley Park Community Consolidated School District No. 146, 864 F.2d 1368 (7th Cir. 1988) (holding that a secretary's delivery of equal pay demands directly to the school board, bypassing the superintendent, was not reasonably proportionate opposition); Hazel v. Postmaster General, 7 F.3d 1 (1st Cir. 1993) (holding that a refusal to work in opposition to perceived discrimination was not reasonably proportionate opposition); EEOC v. Total System Services, Inc., 221 F.3d 1171 (11 th Cir. 2000) (holding that an employer may terminate an employee based on a reasonable, good faith belief that the employee lied in the course of opposing discrimination); Cruz v. Coach Stores, Inc., 202 F.3d 560 (2d Cir. 2000) (holding that an employee's slapping of a sexual harasser is not protected). 2007] Perceiving Subtle Sexism 719 Professor Alex Long has discussed the law's failure to fully protect the challenger's friends and family members from retaliation, leaving a gaping hole in the protection that discrimination law provides against retaliation. 69 A final way in which the discourses and doctrines of discrimination law obstruct perceptions of bias is in the law's emphasis on promoting individuality over group consciousness. Discrimination law's relationship to group consciousness is complex. On the one hand, discrimination law might be thought to promote group consciousness, since it reinforces the salience of social group membership on the basis of race, gender, and other protected classifications. However, the overriding message in recent discrimination jurisprudence extols color-blindness and gender-blindness as the law's animating precepts over an alternative group-based perspective. 7 ° The doctrine treats discrimination as an individual problem, tightly limiting group-based perspectives and relegating them to the margins. Rather than encouraging race or gender consciousness as a way of preventing and remedying discrimination, discrimination law treats it as a necessary evil at best, and tantamount to violating the nondiscrimination principle at worst.'17 Since group consciousness is a key factor in an individual's likelihood of perceiving race or gender bias, as discussed above, the narratives of discrimination law promoting individuality over group-based identification operate to further discourage perceptions of bias. 169 See Alex B. Long, The Troublemaker's Friend: Preventing Retaliation Against Third Parties, Protecting the Right of Association in the Workplace, and Combating Workplace Discrimination, 59 FLA. L. REV. (forthcoming 2007). 170 See, e.g., J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127, 151-54 (1994) (Kennedy, J., concurring) (emphasizing equal protection as an individual right, and insisting that individual jurors decide cases in a gender-neutral fashion); Adarand Constructors, Inc. v. Pena, 515 U.S. 200 (1995) (adopting strict scrutiny for all racial classifications and treating racial classifications burdening minorities as tantamount to racial classifications burdening whites for purposes of selecting a standard of scrutiny). 171 See Reva B. Siegel, Discrimination in the Eyes of the Law: How "Color- Blindness" Discourse Disrupts and Rationalizes Social Stratification, 88 CAL. L. REV. 77 (2000) (discussing and critiquing color-blindness and gender-blindness in antidiscrimination law); Jerome McCrystal Culp, Jr., Colorblindness Remedies and the Intersectionality of Oppression: Policy Arguments Masquerading as Moral Claims, 69 N.Y.U. L. REV. 162, 172, 172 n.38 (1994) (describing and critiquing the Supreme Court's pursuit of color-blindness as a moral objective). 720 Columbia Journal of Gender and Law [Vol. 16:3 III. CONCLUDING THOUGHTS AND AN EXAMPLE: READING STUDENT COURSE EVALUATIONS For an example of how these psychological processes and legal narratives might play out in a particular setting, consider the case of student course evaluations. Much evidence suggests that student course evaluations are influenced by the gender and race of faculty members.172 Women faculty of color and young women faculty, who do not have years of experience to help dispel a presumption of incompetence, may be especially vulnerable to student bias in evaluations.173 Yet law schools treat student evaluations as an objective measure of teaching performance, typically relying on them to evaluate teaching for purposes of tenure, lateral hires, visiting opportunities, promotions, and pay raises. Rarely is their legitimacy and neutrality called into question-at least with regard to their implications for gender and racial equality.1 74 172 See, e.g., Christine M. Bachen et al., Assessing the Role of Gender in College Students' Evaluations of Faculty, 48 CoMM. EDUC. 193 (1999) (describing their research on the influence of sex role expectations on student evaluations of male and female professors and explaining how the gender of students and their adherence to traditional gender schema interacts with the gender of faculty in affecting evaluations); Tamara Baldwin & Nancy Blattner, Guarding Against Potential Bias in Student Evaluations: What Every Faculty Member Needs to Know, 51 C. TEACHING 27, 28 (2003) (citing research identifying "gender and gender-related issues" as sources of bias in student evaluations); Farley, supra note 3, at 334, 336-47 (citing research showing "that student evaluations of women faculty tend to be more hostile than those of male faculty," and discussing anecdotal evidence and results of her study of course evaluations at one law school showing a double-bind facing female law teachers); Rhode, supra note 1, at 482 (acknowledging the double standards and double binds that confront female faculty and citing sources documenting gender bias in the evaluation of female faculty); Deborah Jones Merritt, Bias, the Brain, and Student Evaluations of Teaching, Ohio State Public Law Working Paper No. 87 (2007), available at http://ssm.com/abstract=963196 (explaining the influence of nonverbal behaviors on student evaluations and the process by which race and gender influence evaluations). 173 Cf Chamallas, The Shadow of Professor Kingsfield, supra note 8, at 198 (discussing the legendary Professor Kingsfield from the movie The Paper Chase as the prototype for law professors, and stating that "[s]tudents still expect teachers who look and sound like Kingsfield to be competent, while others have to prove their competency"). 174 See Katherine L. Vaughns, Women of Color in Law Teaching: Shared Identities, Different Experiences, 53 J. LEGAL EDUC. 496, 501 (2003) (noting the potential for bias in student evaluations to negatively affect promotion and retention decisions, and that more credence is given to negative student reactions in such decisions than to positive student observations). A recent article in the Chronicle of Higher Education highlighted evidence that student evaluations are biased by lookism, focusing on the objectification of male faculty, for whom the study found three times as great an effect from looks as it did for women. The article only briefly noted concerns that the evaluations reflect gender and racial Perceiving Subtle Sexism Part of the difficulty in challenging the use of student evaluations is that it is difficult to determine whether sex or race bias has influenced student evaluations in any particular case. Many student comments reviewing women faculty might be interpreted to suggest gender bias, but are subtle or ambiguous. Comments focusing on a female professor's availability, warmth, personality, and helpfulness, for example, might reflect gendered expectations that hold female faculty to higher standards of attentiveness and nurturance, while minimizing and understating intellectual strengths and core competence. 175 Likewise, hostile comments and comments highlighting the professor's political beliefs or values may reflect heightened resistance to female authority, especially a female professor who is or is perceived to be a feminist. On the other hand, at least some of these qualities and characteristics are not in and of themselves irrelevant or illegitimate considerations, if applied evenhandedly and undistorted by gendered expectations. Without blatant and unambiguous signals of prejudice, any individual set of evaluations is unlikely to be perceived as tainted by gender bias. In the culture of the legal academy, in which student evaluations are taken as an indicator of merit, both "just world" ideology and the belief in individual control strongly favor internalizing any fault for a disappointing set of evaluations. For a female faculty member who may be concerned about her evaluations, the tendency toward within-group comparisons also lessens the likelihood that she will discern the presence of bias. If women compare their experiences reading student evaluations with other women, or internally with their own past experience, these comparisons can have a normalizing effect, obscuring any role that gender privilege may play in the process. An added difficulty is the absence of information disclosing across-the-board comparisons, broken down by race, gender, sexual orientation, and other relevant information such as years in teaching, class size, and qualitative differences in written comments. Without aggregate data, cognitive limitations discourage gender-based inferences and favor neutral explanations. bias against women and racial minorities on faculties. Gabriela Montell, Do Good Looks Equal Good Evaluations?, CHRON. HIGHER EDUC., Oct. 15, 2003, http://chronicle.com/jobs/ 2003/10/2003101501c.htm. 175 Farley, supra note 3, at 334 ("When women law professors do receive positive comments, they are much different in nature from the comments received by male professors. Whereas men are most often praised for their 'mastery of the subject matter,' women are usually praised for being enthusiastic and approachable."). 2007] Columbia Journal of Gender and Law The general reluctance on the part of law faculty to raise concerns about bias in student evaluations further contributes to the widespread denial of any problem.1 76 A faculty member who raises concerns about race and gender bias in student evaluations risks being accused of "making excuses" for a particular candidate or not caring about teaching. 7 7 Such a hostile reaction is not surprising, since recognizing flaws in student evaluations would clash with the belief that faculty rewards are based on merit. To the extent that law faculties are steeped in the lessons of discrimination law, we may be less rather than more likely to see race or sex bias in student evaluations as a concern. 7 8 At least absent obvious smoking gun comments revealing prejudice, the storyline of individual fault usually triumphs in discrimination cases. Even if some influence of bias is suspected, discrimination law insists on proof that an overall assessment is negatively affected and translates into actual disadvantage in some material way. Because student evaluations are only one factor among several in an assessment of teaching that may influence a hiring or promotion decision, the ultimate impact of even a clearly biased set of evaluations is far from apparent. By setting up extreme and incontrovertible bias as the paradigm of discrimination, the accumulation of small disadvantages goes unnoticed and unremedied. 1 79 176 Cf Vaughns, supra note 174, at 500 (criticizing the failure of law faculties "to even consider the possibility that students' perceptions of a teacher's ability may be tinged by societal evaluations of racial or gender inferiority, i.e., the 'usual' stereotypes," and their dismissal of critics "as having chips on our shoulders or making much ado about something insignificant"); Farley, supra note 3, at 341 (noting that "many men and women deny the fact that women professors face a more challenging classroom audience; those who complain may even be accused of having a vivid imagination"). 177 Cf Vaughns, supra note 174, at 501 (noting concerns that racial and gender stereotypes influence student evaluations, but stating that "many in academe view student criticisms as legitimate fair assessment of teachers' ability and are unwilling, for whatever reasons, even to acknowledge such a possibility [that they are not]"). Cf Rhode, supra note 1, at 486 (noting that "[w]hen those [gender] issues do arise, students or faculty who express strong views frequently are dismissed or demeaned," and "[w]omen who are open about their feminism, their same-sex orientation, or their views on race have been especially likely targets of offensive comments, adverse student evaluations, and marginalization by their colleagues"). 178 For a general discussion of the failure of discrimination law to address the problems of gender discrimination against women in the academy, see West, supra note 41. 179 See Chamallas, The Shadow of Professor Kingsfield, supra note 8 (discussing the sociological concept of "the accumulation of small disadvantages" and how small blows to reputation and prestige compound and grow over time). [Vol. 16:3722 Perceiving Subtle Sexism Finally, even if we conclude that race or gender bears some share of responsibility for a disappointing set of evaluations, that conclusion does not necessarily protect our self-esteem. More likely, we hold in our mind multiple causes, including self-blame, which wipe out any self-esteem benefit of the perception of bias. As Professor Katherine Vaughns has observed: The insular life of the academic causes the self-doubter to tie her self-esteem to any and all criticisms whether or not justified. And reading student evaluations of her teaching can be a particularly painful experience. As a result, she may internalize every negative judgment expressed in those evaluations despite the presence of positive ones.180 This last point raises a question that may have been nagging at readers (as it has nagged at the author) all along: are we necessarily better off if we heighten our perceptions of gender bias to more clearly see it when it happens? On this score, this Article offers no easy answers. The evidence suggests that perceiving oneself as a target of chronic or systematic bias generally lowers a person's overall sense of well-being. 18' And yet, the failure to perceive bias when it exists has costs too, particularly in reduced opportunities for collective action and opposition.'82 The refusal to name a problem ultimately supports the status quo and closes off opportunities to forge creative strategies for change. For that reason, perhaps recognizing and identifying gender bias, in all its subtle forms, is preferable to denial. 180 Vaughns, supra note 174, at 502-03. 181 Major et al., It's Not My Fault, supra note 116, at 772 (citing studies showing that "perceiving oneself as a target of discrimination is positively associated with depression among women and gay men and with lower self-esteem among women and African Americans") (internal citations omitted); see also id. at 779 (stating that perceiving bias as the cause for a negative event may leave people feeling less depressed than if they perceived an individual attribute such as lack of ability as the cause, but does not lessen the negative effects of other emotions such as hostility and anxiety). 182 See Olson & Hafer, Tolerance of Personal Deprivation, supra note 28, at 167 (arguing that minimizing and denying personal discrimination serves to legitimate the status quo and results in fewer actions to improve one's situation); Jost, Negative Illusions, supra note 28, at 414 (discussing false consciousness and arguing that "errors in cognition produce levels of political acquiescence that may be harmful to the individual, the group, and the society"); see also Magley et al., Outcomes of Self-Labeling, supra note 27, at 399-401 (contending that denying discrimination when it has occurred is not an effective strategy for minimizing its harms). 2007] 723